> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tensor9.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Amazon RDS costs

> Compare Amazon RDS service-adapter costs across backend choices and scale points.

Tensor9 BYOC reference cost models for Amazon RDS. Compare database configurations, storage and availability choices. Matching a price or a vCPU count does not establish equivalent database performance.

Read [how Tensor9 estimates service-adapter costs](/byoc/costs/overview.md) for billing scope, fixed versus variable charges, capacity qualification and a worked example. The comparisons below appear without calculator input.

## Aurora PostgreSQL → AlloyDB for PostgreSQL

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `google::1.0.0::alloydb`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/5468a8f0c9974ac67aab1e746c5c8fdde891d06b78aadf7dcdfa37ac868694ac-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/094e6b229c2844be2147a0e98dca292523626a23e851807f09449d624f8afd28-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 0 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 1460 | 211.7 | AlloyDB for PostgreSQL (one copy) | 191.8805 | -19.8195 |
| 1460 | 211.7 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 142.0653 |
| 2920 | 423.4 | AlloyDB for PostgreSQL (one copy) | 353.7653 | -69.6347 |
| 2920 | 423.4 | AlloyDB for PostgreSQL (two copies) | 677.5349 | 254.1349 |
| 5840 | 846.8 | AlloyDB for PostgreSQL (one copy) | 677.5349 | -169.2651 |
| 5840 | 846.8 | AlloyDB for PostgreSQL (two copies) | 1325.0741 | 478.2741 |
| 11680 | 1693.6 | AlloyDB for PostgreSQL (one copy) | 1325.0741 | -368.5259 |
| 11680 | 1693.6 | AlloyDB for PostgreSQL (two copies) | 2620.1525 | 926.5525 |
| 23360 | 3387.2 | AlloyDB for PostgreSQL (one copy) | 2620.1525 | -767.0475 |
| 23360 | 3387.2 | AlloyDB for PostgreSQL (two copies) | 5210.3093 | 1823.1093 |
| 46720 | 6774.4 | AlloyDB for PostgreSQL (one copy) | 5210.3093 | -1564.0907 |
| 46720 | 6774.4 | AlloyDB for PostgreSQL (two copies) | 10390.6229 | 3616.2229 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* AlloyDB bills each node's vCPU and memory (a high-availability primary is two nodes) and one regional storage pool by used GiB that every instance in the cluster shares, so storage is billed once however many copies run. The smallest machine is c4a-highmem-1 (1 vCPU, 8 GB; [https://docs.cloud.google.com/alloydb/docs/choose-machine-type](https://docs.cloud.google.com/alloydb/docs/choose-machine-type)). C4 and Z3 machine series, backups and committed-use discounts are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* AlloyDB for PostgreSQL pricing: [https://cloud.google.com/alloydb/pricing](https://cloud.google.com/alloydb/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b2d817a7f080d81019ebee2767d317c382dabcef` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [AlloyDB for PostgreSQL pricing](https://cloud.google.com/alloydb/pricing).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/11bd096c069e0e8ac070b85686250dfaf2e68a47b56874a179ff98ebb1931742-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/542573a63b5e92b5d8a92c53d9dc1f81c37d8a7f031ecfaaef39203a54ff5e10-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 0 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 10 | 1 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 190.8805 |
| 10 | 1 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 352.7653 |
| 100 | 10 | AlloyDB for PostgreSQL (one copy) | 223.23035 | 213.23035 |
| 100 | 10 | AlloyDB for PostgreSQL (two copies) | 401.46715 | 391.46715 |
| 1000 | 100 | AlloyDB for PostgreSQL (one copy) | 1971.859575 | 1871.859575 |
| 1000 | 100 | AlloyDB for PostgreSQL (two copies) | 3561.273975 | 3461.273975 |
| 10000 | 1000 | AlloyDB for PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | AlloyDB for PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | AlloyDB for PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | AlloyDB for PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / AlloyDB for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / AlloyDB for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / AlloyDB for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / AlloyDB for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* AlloyDB bills each node's vCPU and memory (a high-availability primary is two nodes) and one regional storage pool by used GiB that every instance in the cluster shares, so storage is billed once however many copies run. The smallest machine is c4a-highmem-1 (1 vCPU, 8 GB; [https://docs.cloud.google.com/alloydb/docs/choose-machine-type](https://docs.cloud.google.com/alloydb/docs/choose-machine-type)). C4 and Z3 machine series, backups and committed-use discounts are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* AlloyDB for PostgreSQL pricing: [https://cloud.google.com/alloydb/pricing](https://cloud.google.com/alloydb/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b2d817a7f080d81019ebee2767d317c382dabcef` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [AlloyDB for PostgreSQL pricing](https://cloud.google.com/alloydb/pricing).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/87c6cfce574cb7f78fffb7cf328010d6643258f32d2b01d4c895fa116c0ec9ce-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/93f5ca6171cd9c16b162e2c7c992801cb94b442161971930c64b084c36d215b8-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 0 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 1 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 1 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 10 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 10 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 100 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 100 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 1000 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 1000 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |
| 10000 | 0 | AlloyDB for PostgreSQL (one copy) | 191.8805 | 191.8805 |
| 10000 | 0 | AlloyDB for PostgreSQL (two copies) | 353.7653 | 353.7653 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / AlloyDB for PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / AlloyDB for PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-month | 0.299957 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| AlloyDB for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.06608 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Target list price · GiB-hour | 0.0112 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Target service minimum | 113.6464 | USD/month |
| AlloyDB for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* AlloyDB bills each node's vCPU and memory (a high-availability primary is two nodes) and one regional storage pool by used GiB that every instance in the cluster shares, so storage is billed once however many copies run. The smallest machine is c4a-highmem-1 (1 vCPU, 8 GB; [https://docs.cloud.google.com/alloydb/docs/choose-machine-type](https://docs.cloud.google.com/alloydb/docs/choose-machine-type)). C4 and Z3 machine series, backups and committed-use discounts are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* AlloyDB for PostgreSQL pricing: [https://cloud.google.com/alloydb/pricing](https://cloud.google.com/alloydb/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b2d817a7f080d81019ebee2767d317c382dabcef` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [AlloyDB for PostgreSQL pricing](https://cloud.google.com/alloydb/pricing).

## Aurora PostgreSQL → AlloyDB for PostgreSQL (RDS proxy)

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `google::1.0.0::alloydb::proxied`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/96ba59415fd774808ccccac3d52cfe6b85766d913be9846f2d0db558f6e5bba6-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/1ae8a64338de373dc5aa0aecb8bc39c5b0857fe67ef74be6e24ad1a3a3f43c83-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1460 | 211.7 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | -86.66 |
| 1460 | 211.7 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 38.38 |
| 2920 | 423.4 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 233.08 | -190.32 |
| 2920 | 423.4 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 466.16 | 42.76 |
| 5840 | 846.8 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 449.16 | -397.64 |
| 5840 | 846.8 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 898.32 | 51.52 |
| 11680 | 1693.6 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 881.32 | -812.28 |
| 11680 | 1693.6 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 1762.64 | 69.04 |
| 23360 | 3387.2 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 1745.64 | -1641.56 |
| 23360 | 3387.2 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 3491.28 | 104.08 |
| 46720 | 6774.4 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 3474.28 | -3300.12 |
| 46720 | 6774.4 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 6948.56 | 174.16 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::alloydb::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-839cb7ac853b15c56ae293dab2a4c10446879765` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/c66687e6d2cf6a6614d8fc56e5405847f0c696e5e8d9b411c78565251583e639-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/958f9fee9dc2c88b1279b64d46a5df39c7a76bc9136be4527e855f4c77c9021f-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10 | 1 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 124.04 |
| 10 | 1 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 249.08 |
| 100 | 10 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 142.3 | 132.3 |
| 100 | 10 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 284.6 | 274.6 |
| 1000 | 100 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 1238.75 | 1138.75 |
| 1000 | 100 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 2477.5 | 2377.5 |
| 10000 | 1000 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / AlloyDB for PostgreSQL (RDS proxy) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::alloydb::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-839cb7ac853b15c56ae293dab2a4c10446879765` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/82c26bfd090a839a99393a66e1eb498793a59d6e2087c8ded0549f53a98449fb-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/8be57cac95e495e2d2720ee1c21cab4b31ff7d9a647bd2d3e3846368c9a40c98-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | AlloyDB for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 1 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 10 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 100 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 100 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1000 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 1000 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10000 | 0 | AlloyDB for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 10000 | 0 | AlloyDB for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / AlloyDB for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / AlloyDB for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| AlloyDB for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::alloydb::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-839cb7ac853b15c56ae293dab2a4c10446879765` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

## Aurora PostgreSQL → Cloud SQL for PostgreSQL

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `google::1.0.0::cloudsql::postgresql`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/c6a704a5ca464e35744f1f620aca14f3e035dc2bd7027cbc66e68a3d327eae61-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/363c7ce1a37f482fd4a50b69af47a128f6432b28ac0ed43496dd7bc6cdc36df4-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 1460 | 211.7 | Cloud SQL for PostgreSQL (one copy) | 118.178 | -93.522 |
| 1460 | 211.7 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 24.656 |
| 2920 | 423.4 | Cloud SQL for PostgreSQL (one copy) | 219.356 | -204.044 |
| 2920 | 423.4 | Cloud SQL for PostgreSQL (two copies) | 438.712 | 15.312 |
| 5840 | 846.8 | Cloud SQL for PostgreSQL (one copy) | 421.712 | -425.088 |
| 5840 | 846.8 | Cloud SQL for PostgreSQL (two copies) | 843.424 | -3.376 |
| 11680 | 1693.6 | Cloud SQL for PostgreSQL (one copy) | 826.424 | -867.176 |
| 11680 | 1693.6 | Cloud SQL for PostgreSQL (two copies) | 1652.848 | -40.752 |
| 23360 | 3387.2 | Cloud SQL for PostgreSQL (one copy) | 1635.848 | -1751.352 |
| 23360 | 3387.2 | Cloud SQL for PostgreSQL (two copies) | 3271.696 | -115.504 |
| 46720 | 6774.4 | Cloud SQL for PostgreSQL (one copy) | 3254.696 | -3519.704 |
| 46720 | 6774.4 | Cloud SQL for PostgreSQL (two copies) | 6509.392 | -265.008 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Cloud SQL Enterprise edition (PostgreSQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-42ce840602ccc44a14430c68c1039d2af3889c41` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/5d286021c00bafedb04572781884ba00fe7abed29813ce08fe89d47619cd029e-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/96fe7c52e703835e03bb3b40e2ffee7f08584141f718a19928160a26f7e8ff15-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 10 | 1 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 117.178 |
| 10 | 1 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 235.356 |
| 100 | 10 | Cloud SQL for PostgreSQL (one copy) | 136.898 | 126.898 |
| 100 | 10 | Cloud SQL for PostgreSQL (two copies) | 273.796 | 263.796 |
| 1000 | 100 | Cloud SQL for PostgreSQL (one copy) | 1210.134 | 1110.134 |
| 1000 | 100 | Cloud SQL for PostgreSQL (two copies) | 2420.268 | 2320.268 |
| 10000 | 1000 | Cloud SQL for PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | Cloud SQL for PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | Cloud SQL for PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | Cloud SQL for PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Cloud SQL for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Cloud SQL for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Cloud SQL for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Cloud SQL for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Cloud SQL Enterprise edition (PostgreSQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-42ce840602ccc44a14430c68c1039d2af3889c41` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/a3cf241a3f1a15535c9174e811511e8cc45ffc911eee97b19d689e9e5d72929f-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/b34041d4d9b3d402e51b4a0997417344914ac7894c6ed11e04da3196396521c6-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 1 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 1 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 10 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 10 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 100 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 100 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 1000 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 1000 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |
| 10000 | 0 | Cloud SQL for PostgreSQL (one copy) | 118.178 | 118.178 |
| 10000 | 0 | Cloud SQL for PostgreSQL (two copies) | 236.356 | 236.356 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Cloud SQL Enterprise edition (PostgreSQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-42ce840602ccc44a14430c68c1039d2af3889c41` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

## Aurora PostgreSQL → OCI Database with PostgreSQL

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `oci::1.0.0::psql`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/16dd0bb1c820d515e35fd25d2054d8274d611cf539d8fb4438a626f951256b28-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/abe696731f732b81827079892b089b01bffb23fb9870cdec6cf605ee765b9f8c-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1460 | 211.7 | OCI Database with PostgreSQL (one copy) | 94.9 | -116.8 |
| 1460 | 211.7 | OCI Database with PostgreSQL (two copies) | 173.64 | -38.06 |
| 2920 | 423.4 | OCI Database with PostgreSQL (one copy) | 173.64 | -249.76 |
| 2920 | 423.4 | OCI Database with PostgreSQL (two copies) | 340.08 | -83.32 |
| 5840 | 846.8 | OCI Database with PostgreSQL (one copy) | 340.08 | -506.72 |
| 5840 | 846.8 | OCI Database with PostgreSQL (two copies) | 672.96 | -173.84 |
| 11680 | 1693.6 | OCI Database with PostgreSQL (one copy) | 672.96 | -1020.64 |
| 11680 | 1693.6 | OCI Database with PostgreSQL (two copies) | 1338.72 | -354.88 |
| 23360 | 3387.2 | OCI Database with PostgreSQL (one copy) | 1338.72 | -2048.48 |
| 23360 | 3387.2 | OCI Database with PostgreSQL (two copies) | 2670.24 | -716.96 |
| 46720 | 6774.4 | OCI Database with PostgreSQL (one copy) | 2670.24 | -4104.16 |
| 46720 | 6774.4 | OCI Database with PostgreSQL (two copies) | 5333.28 | -1441.12 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9631e4e7ad853051d4ac9c6ec36873c05803b586` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/b7c738ba7e41b09695761c110f0d052011b70d18572a504a221c74852eb89fd5-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/7af3be766c7aaa4fcae7c565fe12d9ee7504aa046741e5ee45ee7a345c34f97f-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10 | 1 | OCI Database with PostgreSQL (one copy) | 94.9 | 93.9 |
| 10 | 1 | OCI Database with PostgreSQL (two copies) | 173.64 | 172.64 |
| 100 | 10 | OCI Database with PostgreSQL (one copy) | 96.94 | 86.94 |
| 100 | 10 | OCI Database with PostgreSQL (two copies) | 183.08 | 173.08 |
| 1000 | 100 | OCI Database with PostgreSQL (one copy) | 810.12 | 710.12 |
| 1000 | 100 | OCI Database with PostgreSQL (two copies) | 1528.44 | 1428.44 |
| 10000 | 1000 | OCI Database with PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | OCI Database with PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | OCI Database with PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | OCI Database with PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / OCI Database with PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / OCI Database with PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / OCI Database with PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / OCI Database with PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9631e4e7ad853051d4ac9c6ec36873c05803b586` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/079b18c999255bc3d220118a7ef06e67050f2cde566484a06e5142c5c7442d0b-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/7b8ca1fd12fed7e27cd2eb16e1b4b042efbce430d3ee3c74a39a0f7416555686-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 1 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 10 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 100 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 100 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1000 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 1000 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10000 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 10000 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9631e4e7ad853051d4ac9c6ec36873c05803b586` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

## Aurora PostgreSQL → PostgreSQL (CloudNativePG)

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `cloudnative::1.0.0::postgresql`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/502531d47a41072763be6bdc22f5b65503c533998f501e5f771ec83aa553b677-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/2d5d126fb015bf560632a245501d2c2083f2ba08100dc040959ef0b0cf3a19d4-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1460 | 211.7 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | -130.80386 |
| 1460 | 211.7 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | -49.90772 |
| 2920 | 423.4 | PostgreSQL (CloudNativePG) (one copy) | 151.79228 | -271.60772 |
| 2920 | 423.4 | PostgreSQL (CloudNativePG) (two copies) | 303.58456 | -119.81544 |
| 5840 | 846.8 | PostgreSQL (CloudNativePG) (one copy) | 293.58456 | -553.21544 |
| 5840 | 846.8 | PostgreSQL (CloudNativePG) (two copies) | 587.16912 | -259.63088 |
| 11680 | 1693.6 | PostgreSQL (CloudNativePG) (one copy) | 577.16912 | -1116.43088 |
| 11680 | 1693.6 | PostgreSQL (CloudNativePG) (two copies) | 1154.33824 | -539.26176 |
| 23360 | 3387.2 | PostgreSQL (CloudNativePG) (one copy) | 1144.33824 | -2242.86176 |
| 23360 | 3387.2 | PostgreSQL (CloudNativePG) (two copies) | 2288.67648 | -1098.52352 |
| 46720 | 6774.4 | PostgreSQL (CloudNativePG) (one copy) | 2278.67648 | -4495.72352 |
| 46720 | 6774.4 | PostgreSQL (CloudNativePG) (two copies) | 4557.35296 | -2217.04704 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-dd3a8615165abfe7fa17fb70889dc6979bc55d1a` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/70e8a94458ab54cb20188b6fe2e248e8dda5270fc25e309ae9dceedf46ab8097-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/c735ab7bef3ca3b459a9a108518acdd33c2ab9c55040f8d65352ebd801646fe9-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10 | 1 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 79.89614 |
| 10 | 1 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 160.79228 |
| 100 | 10 | PostgreSQL (CloudNativePG) (one copy) | 92.08216 | 82.08216 |
| 100 | 10 | PostgreSQL (CloudNativePG) (two copies) | 184.16432 | 174.16432 |
| 1000 | 100 | PostgreSQL (CloudNativePG) (one copy) | 806.01748 | 706.01748 |
| 1000 | 100 | PostgreSQL (CloudNativePG) (two copies) | 1612.03496 | 1512.03496 |
| 10000 | 1000 | PostgreSQL (CloudNativePG) (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | PostgreSQL (CloudNativePG) (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | PostgreSQL (CloudNativePG) (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | PostgreSQL (CloudNativePG) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-dd3a8615165abfe7fa17fb70889dc6979bc55d1a` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/8382b3b78e6735c7faaf03fe23c78d86c3d350d7c9c48c5082b0770089df9df6-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/a82c811fd80e1b60974fc186700774c52057cc118cfab7a50e6158bc1d7f6773-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 1 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 10 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 100 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 100 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1000 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 1000 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10000 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 10000 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-dd3a8615165abfe7fa17fb70889dc6979bc55d1a` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

## Aurora PostgreSQL → PostgreSQL Flexible Server

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `azure::1.0.0::postgresql::flexible`.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/cd01d9d987adfa3020d118476dfd884114a2e774c2830703252b19477c62fd48-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/9b45dba973a6e5289bdf0373a08aaaf1be424bf70402cf3e843c98c87db91533-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1460 | 211.7 | PostgreSQL Flexible Server (one copy) | 141.44 | -70.26 |
| 1460 | 211.7 | PostgreSQL Flexible Server (two copies) | 282.88 | 71.18 |
| 2920 | 423.4 | PostgreSQL Flexible Server (one copy) | 271.38 | -152.02 |
| 2920 | 423.4 | PostgreSQL Flexible Server (two copies) | 542.76 | 119.36 |
| 5840 | 846.8 | PostgreSQL Flexible Server (one copy) | 531.26 | -315.54 |
| 5840 | 846.8 | PostgreSQL Flexible Server (two copies) | 1062.52 | 215.72 |
| 11680 | 1693.6 | PostgreSQL Flexible Server (one copy) | 1051.02 | -642.58 |
| 11680 | 1693.6 | PostgreSQL Flexible Server (two copies) | 2102.04 | 408.44 |
| 23360 | 3387.2 | PostgreSQL Flexible Server (one copy) | 2090.54 | -1296.66 |
| 23360 | 3387.2 | PostgreSQL Flexible Server (two copies) | 4181.08 | 793.88 |
| 46720 | 6774.4 | PostgreSQL Flexible Server (one copy) | 4169.58 | -2604.82 |
| 46720 | 6774.4 | PostgreSQL Flexible Server (two copies) | 8339.16 | 1564.76 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-872daf3fc345e314056ab6183e28b95413202233` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/3340e58da3878a6fe1fb20410e3c7d956878d89a89b1182cc1666ee120c7c12e-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/47e9273203f04f02fa84f2b6061c7a1791ccce2ea4ddf34006c5f7cb7e0d65be-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10 | 1 | PostgreSQL Flexible Server (one copy) | 141.44 | 140.44 |
| 10 | 1 | PostgreSQL Flexible Server (two copies) | 282.88 | 281.88 |
| 100 | 10 | PostgreSQL Flexible Server (one copy) | 160.33 | 150.33 |
| 100 | 10 | PostgreSQL Flexible Server (two copies) | 320.66 | 310.66 |
| 1000 | 100 | PostgreSQL Flexible Server (one copy) | 1422.665 | 1322.665 |
| 1000 | 100 | PostgreSQL Flexible Server (two copies) | 2845.33 | 2745.33 |
| 10000 | 1000 | PostgreSQL Flexible Server (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | PostgreSQL Flexible Server (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | PostgreSQL Flexible Server (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | PostgreSQL Flexible Server (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / PostgreSQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / PostgreSQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / PostgreSQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / PostgreSQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-872daf3fc345e314056ab6183e28b95413202233` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/5febfb290ca6287e1b87b9ad0db8b7988499483292c13179db1e58726244bbe6-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/c0856a8ed89f94f6ca6a8083c602459ecf1d2270c1fa2c234d8dd4399f924a55-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 1 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 10 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 100 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 100 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1000 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 1000 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10000 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 10000 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-872daf3fc345e314056ab6183e28b95413202233` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

## Aurora PostgreSQL → Scaleway Managed Database for PostgreSQL

Origin API: `aws::1.0.0::rds::aurora-postgresql`. Target backend: `scaleway::1.0.0::rdb::postgresql`.

**Native or deployment mapping.** This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA. Cost estimates describe the selected service resources; operation coverage remains separate.

### Aurora PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/1188c0933bcb47a812eb1d80bb786ee8c6d8b10363270222ab7913003cd0ff4b-light.svg)

![Aurora PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/a8b8fcf671193c48e7cce030a308f536ef812bf926fd148e90097018e14d6dff-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | Aurora PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1460 | 211.7 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | -110.597285 |
| 1460 | 211.7 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | -57.01284605 |
| 2920 | 423.4 | Scaleway Managed Database for PostgreSQL (one copy) | 191.078865 | -232.321135 |
| 2920 | 423.4 | Scaleway Managed Database for PostgreSQL (two copies) | 292.35066345 | -131.04933655 |
| 5840 | 846.8 | Scaleway Managed Database for PostgreSQL (one copy) | 371.031165 | -475.768835 |
| 5840 | 846.8 | Scaleway Managed Database for PostgreSQL (two copies) | 567.67768245 | -279.12231755 |
| 11680 | 1693.6 | Scaleway Managed Database for PostgreSQL (one copy) | 730.935765 | -962.664235 |
| 11680 | 1693.6 | Scaleway Managed Database for PostgreSQL (two copies) | 1118.33172045 | -575.26827955 |
| 23360 | 3387.2 | Scaleway Managed Database for PostgreSQL (one copy) | 1450.744965 | -1936.455035 |
| 23360 | 3387.2 | Scaleway Managed Database for PostgreSQL (two copies) | 2219.63979645 | -1167.56020355 |
| 46720 | 6774.4 | Scaleway Managed Database for PostgreSQL (one copy) | 2890.363365 | -3884.036635 |
| 46720 | 6774.4 | Scaleway Managed Database for PostgreSQL (two copies) | 4422.25594845 | -2352.14405155 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9818fc9ac977d4349d19fe5ce8eb48812af182c2` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### Aurora PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/1267443a540f020842d85ee1ce2abacc5329becad8b5047ea49b9b03ba587ff9-light.svg)

![Aurora PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/dfef67a683b9e64217d38b9f2facde97d758d24bb786065e988389f81368663b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | Aurora PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 1 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 100.102715 |
| 10 | 1 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 153.68715395 |
| 100 | 10 | Scaleway Managed Database for PostgreSQL (one copy) | 115.7544975 | 105.7544975 |
| 100 | 10 | Scaleway Managed Database for PostgreSQL (two copies) | 177.104381175 | 167.104381175 |
| 1000 | 100 | Scaleway Managed Database for PostgreSQL (one copy) | 1025.26590375 | 925.26590375 |
| 1000 | 100 | Scaleway Managed Database for PostgreSQL (two copies) | 1568.6568327375 | 1468.6568327375 |
| 10000 | 1000 | Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1000 | Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 10000 | Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 10000 | Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9818fc9ac977d4349d19fe5ce8eb48812af182c2` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### Aurora PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/d7133f9871f0cff14d68aaa93ebb8ec746406793d7afdf602d03cdd410404cd8-light.svg)

![Aurora PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/1730165b9c5bf78cb678df3ff170c4c64140791f743a22c874f931b7531e28b7-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | Aurora PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 1 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 10 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 100 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 100 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1000 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 1000 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10000 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 10000 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.145 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.1 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* Aurora Standard also bills storage I/O at USD 0.20 per 1M I/Os (buffer-cache misses, not logical operations); no meter prices it, so the origin cost excludes it. I/O-Optimized clusters bill no I/O but 30% more per instance and 0.225 per GB-month storage.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-9818fc9ac977d4349d19fe5ce8eb48812af182c2` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

## RDS MySQL → Cloud SQL for MySQL

Origin API: `aws::1.0.0::rds::mysql`. Target backend: `google::1.0.0::cloudsql::mysql`.

### RDS MySQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/42111cbe3a95f22ea9893585dde64075a52c1b795c1227a49fb4edbcf641bb57-light.svg)

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/a11bb55a6b4dbadef957f6906578fc85b6d89b25a2baea972459b49fb0d13ab6-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS MySQL | Target scenario | Cloud SQL for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 1460 | 124.83 | Cloud SQL for MySQL (one copy) | 118.178 | -6.652 |
| 1460 | 124.83 | Cloud SQL for MySQL (two copies) | 236.356 | 111.526 |
| 2920 | 249.66 | Cloud SQL for MySQL (one copy) | 219.356 | -30.304 |
| 2920 | 249.66 | Cloud SQL for MySQL (two copies) | 438.712 | 189.052 |
| 5840 | 499.32 | Cloud SQL for MySQL (one copy) | 421.712 | -77.608 |
| 5840 | 499.32 | Cloud SQL for MySQL (two copies) | 843.424 | 344.104 |
| 11680 | 998.64 | Cloud SQL for MySQL (one copy) | 826.424 | -172.216 |
| 11680 | 998.64 | Cloud SQL for MySQL (two copies) | 1652.848 | 654.208 |
| 23360 | 1997.28 | Cloud SQL for MySQL (one copy) | 1635.848 | -361.432 |
| 23360 | 1997.28 | Cloud SQL for MySQL (two copies) | 3271.696 | 1274.416 |
| 46720 | 3994.56 | Cloud SQL for MySQL (one copy) | 3254.696 | -739.864 |
| 46720 | 3994.56 | Cloud SQL for MySQL (two copies) | 6509.392 | 2514.832 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Cloud SQL Enterprise edition (MySQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ee670329310268dd09cbb5242012e35e37c2b109` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

### RDS MySQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/ee6276b22b3393b90f712ed64b4fa16eb9545a19f2e4930a777f4d859492f022-light.svg)

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/6b557ed683447698cf80023172131bfcdc2be29d76ffb2daede23bac8975403e-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS MySQL | Target scenario | Cloud SQL for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 10 | 1.15 | Cloud SQL for MySQL (one copy) | 118.178 | 117.028 |
| 10 | 1.15 | Cloud SQL for MySQL (two copies) | 236.356 | 235.206 |
| 100 | 11.5 | Cloud SQL for MySQL (one copy) | 136.898 | 125.398 |
| 100 | 11.5 | Cloud SQL for MySQL (two copies) | 273.796 | 262.296 |
| 1000 | 115 | Cloud SQL for MySQL (one copy) | 1210.134 | 1095.134 |
| 1000 | 115 | Cloud SQL for MySQL (two copies) | 2420.268 | 2305.268 |
| 10000 | 1150 | Cloud SQL for MySQL (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | Cloud SQL for MySQL (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | Cloud SQL for MySQL (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | Cloud SQL for MySQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Cloud SQL for MySQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Cloud SQL for MySQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Cloud SQL for MySQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Cloud SQL for MySQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Cloud SQL Enterprise edition (MySQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ee670329310268dd09cbb5242012e35e37c2b109` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

### RDS MySQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/f71d7f8e40a0896828d1b76be6246cb5ddb71fad6f5c2b1cf0b91b5670027f51-light.svg)

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/d62869e928a54d3068de10c20c14cb3f45046cff49116e08a51520f0e058a465-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS MySQL | Target scenario | Cloud SQL for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 0 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 1 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 1 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 10 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 10 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 100 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 100 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 1000 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 1000 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |
| 10000 | 0 | Cloud SQL for MySQL (one copy) | 118.178 | 118.178 |
| 10000 | 0 | Cloud SQL for MySQL (two copies) | 236.356 | 236.356 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for MySQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for MySQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (one copy) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (one copy) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (one copy) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for MySQL (two copies) | Target list price · network GiB | 0.19 | USD/network GiB |
| Cloud SQL for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for MySQL (two copies) | Target list price · vCPU-hour | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Target list price · GiB-hour | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for MySQL (two copies) | Target service minimum | 9.365 | USD/month |
| Cloud SQL for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Cloud SQL Enterprise edition (MySQL) publishes separate vCPU, memory and SSD prices for zonal instances; a high-availability (regional) instance bills exactly twice each, which the standby copy reproduces. Enterprise Plus, committed-use discounts, backups (USD 0.08 per GiB-month) and idle IPv4 addresses (0.01 per hour) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL pricing: [https://cloud.google.com/sql/pricing](https://cloud.google.com/sql/pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ee670329310268dd09cbb5242012e35e37c2b109` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL pricing](https://cloud.google.com/sql/pricing).

## RDS MySQL → MySQL Flexible Server

Origin API: `aws::1.0.0::rds::mysql`. Target backend: `azure::1.0.0::mysql::flexible`.

### RDS MySQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/4c51855963ca96f9e21193787f167e8886618bdb8684ffa901875d453dab4e10-light.svg)

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/22b063c392e42b70ad6b869ca7868212765910321167113aed259c4ccab86fbd-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS MySQL | Target scenario | MySQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 0 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 1460 | 124.83 | MySQL Flexible Server (one copy) | 136.33 | 11.5 |
| 1460 | 124.83 | MySQL Flexible Server (two copies) | 272.66 | 147.83 |
| 2920 | 249.66 | MySQL Flexible Server (one copy) | 261.16 | 11.5 |
| 2920 | 249.66 | MySQL Flexible Server (two copies) | 522.32 | 272.66 |
| 5840 | 499.32 | MySQL Flexible Server (one copy) | 510.82 | 11.5 |
| 5840 | 499.32 | MySQL Flexible Server (two copies) | 1021.64 | 522.32 |
| 11680 | 998.64 | MySQL Flexible Server (one copy) | 1010.14 | 11.5 |
| 11680 | 998.64 | MySQL Flexible Server (two copies) | 2020.28 | 1021.64 |
| 23360 | 1997.28 | MySQL Flexible Server (one copy) | 2008.78 | 11.5 |
| 23360 | 1997.28 | MySQL Flexible Server (two copies) | 4017.56 | 2020.28 |
| 46720 | 3994.56 | MySQL Flexible Server (one copy) | 4006.06 | 11.5 |
| 46720 | 3994.56 | MySQL Flexible Server (two copies) | 8012.12 | 4017.56 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| MySQL Flexible Server (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (one copy) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| MySQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| MySQL Flexible Server (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (two copies) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| MySQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for MySQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 127.13 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-a14b5a17bd919ad2be9588e5e7a7809815efbb8e` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### RDS MySQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/dd84b8e374441d16ffe2f6e29677b07734b8d20382f6e60dad3e9a9b71417d47-light.svg)

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/bb69f23c596549cf723a51a47651783320df5c16d63d276621b4c45c1ff7d1c9-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS MySQL | Target scenario | MySQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 0 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 10 | 1.15 | MySQL Flexible Server (one copy) | 136.33 | 135.18 |
| 10 | 1.15 | MySQL Flexible Server (two copies) | 272.66 | 271.51 |
| 100 | 11.5 | MySQL Flexible Server (one copy) | 153.9425 | 142.4425 |
| 100 | 11.5 | MySQL Flexible Server (two copies) | 307.885 | 296.385 |
| 1000 | 115 | MySQL Flexible Server (one copy) | 1358.79 | 1243.79 |
| 1000 | 115 | MySQL Flexible Server (two copies) | 2717.58 | 2602.58 |
| 10000 | 1150 | MySQL Flexible Server (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | MySQL Flexible Server (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | MySQL Flexible Server (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | MySQL Flexible Server (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / MySQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / MySQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / MySQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / MySQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| MySQL Flexible Server (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (one copy) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| MySQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| MySQL Flexible Server (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (two copies) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| MySQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for MySQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 127.13 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-a14b5a17bd919ad2be9588e5e7a7809815efbb8e` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### RDS MySQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/f90e1336ca9c73ca2e8ee60d32afe3dd3074dd8402c6eeb9da27f8c9daed9880-light.svg)

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/1e5d939f2b7a79d1469ca7ee0fed31b9330879f19ea83e70430ce1a17388adc7-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS MySQL | Target scenario | MySQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 0 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 1 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 1 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 10 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 10 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 100 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 100 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 1000 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 1000 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |
| 10000 | 0 | MySQL Flexible Server (one copy) | 136.33 | 136.33 |
| 10000 | 0 | MySQL Flexible Server (two copies) | 272.66 | 272.66 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / MySQL Flexible Server (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / MySQL Flexible Server (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| MySQL Flexible Server (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (one copy) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| MySQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| MySQL Flexible Server (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| MySQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| MySQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| MySQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| MySQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| MySQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| MySQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| MySQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| MySQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| MySQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| MySQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Target list price · GiB-hour | 0.008125 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| MySQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| MySQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| MySQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| MySQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| MySQL Flexible Server (two copies) | Target service minimum | 8.505 | USD/month |
| MySQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| MySQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| MySQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| MySQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| MySQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| MySQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| MySQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| MySQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| MySQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| MySQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| MySQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| MySQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| MySQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| MySQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| MySQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for MySQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 127.13 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-a14b5a17bd919ad2be9588e5e7a7809815efbb8e` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

## RDS MySQL → OCI MySQL Database (HeatWave)

Origin API: `aws::1.0.0::rds::mysql`. Target backend: `oci::1.0.0::mysql`.

### RDS MySQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/134fa04bf1324ebdd08af28e362e487ef2d6fad8e7c96f383c092dfe21a46dcf-light.svg)

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/21b6e8e901d342fa7285d104d3dcdf417f44c1bf8a23148475fce16c72cfe62a-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS MySQL | Target scenario | OCI MySQL Database (HeatWave) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 0 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 1460 | 124.83 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | -67.394 |
| 1460 | 124.83 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 47.478 |
| 2920 | 249.66 | OCI MySQL Database (HeatWave) (one copy) | 110.872 | -138.788 |
| 2920 | 249.66 | OCI MySQL Database (HeatWave) (two copies) | 332.616 | 82.956 |
| 5840 | 499.32 | OCI MySQL Database (HeatWave) (one copy) | 217.744 | -281.576 |
| 5840 | 499.32 | OCI MySQL Database (HeatWave) (two copies) | 653.232 | 153.912 |
| 11680 | 998.64 | OCI MySQL Database (HeatWave) (one copy) | 431.488 | -567.152 |
| 11680 | 998.64 | OCI MySQL Database (HeatWave) (two copies) | 1294.464 | 295.824 |
| 23360 | 1997.28 | OCI MySQL Database (HeatWave) (one copy) | 858.976 | -1138.304 |
| 23360 | 1997.28 | OCI MySQL Database (HeatWave) (two copies) | 2576.928 | 579.648 |
| 46720 | 3994.56 | OCI MySQL Database (HeatWave) (one copy) | 1713.952 | -2280.608 |
| 46720 | 3994.56 | OCI MySQL Database (HeatWave) (two copies) | 5141.856 | 1147.296 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (one copy) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI MySQL Database (HeatWave) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (two copies) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI MySQL Database (HeatWave) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI HeatWave MySQL bills ECPUs per hour with 8 GiB of memory included per ECPU (MySQL.2 is 2 ECPU and 16 GiB; [https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html)), so memory has no separate rate and one ECPU is treated as one vCPU; a sizing above 8 GiB per vCPU needs a larger shape this rate does not add. A high-availability DB system runs three instances ([https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html)), so a standby selection adds two copies of compute and storage. HeatWave clusters, backups and the MySQL.Free shape are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ca5859d924b1dc2b224b756d50f256b36b5e6d19` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/).

### RDS MySQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/4dae12258b8700d3330940b2600e0569eff871396779f02dfe49399150e1fc3d-light.svg)

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/7686d66f4f8392d59207ff46fa1874941ed257117ed1e9e0476d9b643f583380-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS MySQL | Target scenario | OCI MySQL Database (HeatWave) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 0 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 10 | 1.15 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 56.286 |
| 10 | 1.15 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 171.158 |
| 100 | 11.5 | OCI MySQL Database (HeatWave) (one copy) | 59.436 | 47.936 |
| 100 | 11.5 | OCI MySQL Database (HeatWave) (two copies) | 178.308 | 166.808 |
| 1000 | 115 | OCI MySQL Database (HeatWave) (one copy) | 478.488 | 363.488 |
| 1000 | 115 | OCI MySQL Database (HeatWave) (two copies) | 1435.464 | 1320.464 |
| 10000 | 1150 | OCI MySQL Database (HeatWave) (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | OCI MySQL Database (HeatWave) (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | OCI MySQL Database (HeatWave) (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | OCI MySQL Database (HeatWave) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / OCI MySQL Database (HeatWave) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / OCI MySQL Database (HeatWave) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / OCI MySQL Database (HeatWave) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / OCI MySQL Database (HeatWave) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (one copy) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI MySQL Database (HeatWave) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (two copies) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI MySQL Database (HeatWave) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI HeatWave MySQL bills ECPUs per hour with 8 GiB of memory included per ECPU (MySQL.2 is 2 ECPU and 16 GiB; [https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html)), so memory has no separate rate and one ECPU is treated as one vCPU; a sizing above 8 GiB per vCPU needs a larger shape this rate does not add. A high-availability DB system runs three instances ([https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html)), so a standby selection adds two copies of compute and storage. HeatWave clusters, backups and the MySQL.Free shape are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ca5859d924b1dc2b224b756d50f256b36b5e6d19` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/).

### RDS MySQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/b9f1eac3ba1b3285356648e0be37ad870170e793e96afc8c7204592a5a9ce997-light.svg)

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/367bb8b07e76db6d53a3ecee351d4a3dc928e94c51da1a02a1b0f3266de8344b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS MySQL | Target scenario | OCI MySQL Database (HeatWave) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 0 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 1 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 1 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 10 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 10 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 100 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 100 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 1000 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 1000 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |
| 10000 | 0 | OCI MySQL Database (HeatWave) (one copy) | 57.436 | 57.436 |
| 10000 | 0 | OCI MySQL Database (HeatWave) (two copies) | 172.308 | 172.308 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI MySQL Database (HeatWave) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI MySQL Database (HeatWave) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (one copy) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI MySQL Database (HeatWave) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI MySQL Database (HeatWave) (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target disk / source disk | 1 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · read | 0 | USD/read |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · write | 0 | USD/write |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · GiB-month | 0.04 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Target list price · vCPU-hour | 0.0366 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · GiB-hour | 0 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI MySQL Database (HeatWave) (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI MySQL Database (HeatWave) (two copies) | Target monthly base charge | 0 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Target service minimum | 53.436 | USD/month |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary RAM | 8 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Minimum primary disk | 100 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI MySQL Database (HeatWave) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI MySQL Database (HeatWave) (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI MySQL Database (HeatWave) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI MySQL Database (HeatWave) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI MySQL Database (HeatWave) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI MySQL Database (HeatWave) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI MySQL Database (HeatWave) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI HeatWave MySQL bills ECPUs per hour with 8 GiB of memory included per ECPU (MySQL.2 is 2 ECPU and 16 GiB; [https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/supported-shapes.html)), so memory has no separate rate and one ECPU is treated as one vCPU; a sizing above 8 GiB per vCPU needs a larger shape this rate does not add. A high-availability DB system runs three instances ([https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html](https://docs.oracle.com/en-us/iaas/mysql-database/doc/overview-high-availability.html)), so a standby selection adds two copies of compute and storage. HeatWave clusters, backups and the MySQL.Free shape are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-ca5859d924b1dc2b224b756d50f256b36b5e6d19` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/).

## RDS MySQL → Percona XtraDB Cluster (MySQL)

Origin API: `aws::1.0.0::rds::mysql`. Target backend: `percona::1.0.0::mysql`.

### RDS MySQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/6353045e545bd86bd05e1844b71c32bca9816eb4fb5315eeebeeeaa896f12de2-light.svg)

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/f03a1803c9f1d5eecf02bb7d545a264895b72a85c785b2d33dddb1fc900bd570-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS MySQL | Target scenario | Percona XtraDB Cluster (MySQL) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 1460 | 124.83 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | -43.93386 |
| 1460 | 124.83 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 36.96228 |
| 2920 | 249.66 | Percona XtraDB Cluster (MySQL) (one copy) | 151.79228 | -97.86772 |
| 2920 | 249.66 | Percona XtraDB Cluster (MySQL) (two copies) | 303.58456 | 53.92456 |
| 5840 | 499.32 | Percona XtraDB Cluster (MySQL) (one copy) | 293.58456 | -205.73544 |
| 5840 | 499.32 | Percona XtraDB Cluster (MySQL) (two copies) | 587.16912 | 87.84912 |
| 11680 | 998.64 | Percona XtraDB Cluster (MySQL) (one copy) | 577.16912 | -421.47088 |
| 11680 | 998.64 | Percona XtraDB Cluster (MySQL) (two copies) | 1154.33824 | 155.69824 |
| 23360 | 1997.28 | Percona XtraDB Cluster (MySQL) (one copy) | 1144.33824 | -852.94176 |
| 23360 | 1997.28 | Percona XtraDB Cluster (MySQL) (two copies) | 2288.67648 | 291.39648 |
| 46720 | 3994.56 | Percona XtraDB Cluster (MySQL) (one copy) | 2278.67648 | -1715.88352 |
| 46720 | 3994.56 | Percona XtraDB Cluster (MySQL) (two copies) | 4557.35296 | 562.79296 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: Percona Operator for MySQL (Apache-2.0, [https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE](https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE)) with Percona Server for MySQL (GPLv2). Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-86c8dfa5bdc4a23be3e057a087872f347e864a44` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### RDS MySQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/310841e110ae5694e85e0c47b585b4f0e13ed6a427fb309a13ff69b91b1b210f-light.svg)

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/72b9b89a8354a92dc3e7ff7e0c7d3a3fe779a9ef0c81895294a26f4162b86d87-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS MySQL | Target scenario | Percona XtraDB Cluster (MySQL) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 10 | 1.15 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 79.74614 |
| 10 | 1.15 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 160.64228 |
| 100 | 11.5 | Percona XtraDB Cluster (MySQL) (one copy) | 92.08216 | 80.58216 |
| 100 | 11.5 | Percona XtraDB Cluster (MySQL) (two copies) | 184.16432 | 172.66432 |
| 1000 | 115 | Percona XtraDB Cluster (MySQL) (one copy) | 806.01748 | 691.01748 |
| 1000 | 115 | Percona XtraDB Cluster (MySQL) (two copies) | 1612.03496 | 1497.03496 |
| 10000 | 1150 | Percona XtraDB Cluster (MySQL) (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | Percona XtraDB Cluster (MySQL) (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | Percona XtraDB Cluster (MySQL) (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | Percona XtraDB Cluster (MySQL) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Percona XtraDB Cluster (MySQL) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Percona XtraDB Cluster (MySQL) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: Percona Operator for MySQL (Apache-2.0, [https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE](https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE)) with Percona Server for MySQL (GPLv2). Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-86c8dfa5bdc4a23be3e057a087872f347e864a44` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### RDS MySQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/9aa5f0d919e36c41a7fe1d7571654d55830d385e1f56946cda03636588ce8662-light.svg)

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/764eca682d04f6e4e6bd7ae7fb268e8b79167c77564c551d0d1c35c349cd697b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS MySQL | Target scenario | Percona XtraDB Cluster (MySQL) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 1 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 1 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 10 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 10 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 100 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 100 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 1000 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 1000 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |
| 10000 | 0 | Percona XtraDB Cluster (MySQL) (one copy) | 80.89614 | 80.89614 |
| 10000 | 0 | Percona XtraDB Cluster (MySQL) (two copies) | 161.79228 | 161.79228 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Percona XtraDB Cluster (MySQL) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Percona XtraDB Cluster (MySQL) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (one copy) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Percona XtraDB Cluster (MySQL) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target disk / source disk | 1 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · read | 0 | USD/read |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · write | 0 | USD/write |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · input token | 0 | USD/input token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target price guess · output token | 0 | USD/output token |
| Percona XtraDB Cluster (MySQL) (two copies) | Target monthly base charge | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Approximate target service minimum | 0 | USD/month |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary CPU | 2 | vCPU |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary RAM | 8 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Minimum primary disk | 100 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Percona XtraDB Cluster (MySQL) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning peak / mean work | 2 | multiplier |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Planning disk utilization | 0.8 | fraction |
| Percona XtraDB Cluster (MySQL) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Percona XtraDB Cluster (MySQL) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Percona XtraDB Cluster (MySQL) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: Percona Operator for MySQL (Apache-2.0, [https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE](https://github.com/percona/percona-server-mysql-operator/blob/main/LICENSE)) with Percona Server for MySQL (GPLv2). Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-86c8dfa5bdc4a23be3e057a087872f347e864a44` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

## RDS MySQL → Scaleway Managed Database for MySQL

Origin API: `aws::1.0.0::rds::mysql`. Target backend: `scaleway::1.0.0::rdb::mysql`.

**Native or deployment mapping.** This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA. Cost estimates describe the selected service resources; operation coverage remains separate.

### RDS MySQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/5c77321c6aea798eb7bf16d23cad211a5185e11c44889735d91ef979278a85ba-light.svg)

![RDS MySQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/b81c8832870d5580c7941448f9ef1408e9b35e490448984903a49a93072bfaa1-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS MySQL | Target scenario | Scaleway Managed Database for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 1460 | 124.83 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | -23.727285 |
| 1460 | 124.83 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 29.85715395 |
| 2920 | 249.66 | Scaleway Managed Database for MySQL (one copy) | 191.078865 | -58.581135 |
| 2920 | 249.66 | Scaleway Managed Database for MySQL (two copies) | 292.35066345 | 42.69066345 |
| 5840 | 499.32 | Scaleway Managed Database for MySQL (one copy) | 371.031165 | -128.288835 |
| 5840 | 499.32 | Scaleway Managed Database for MySQL (two copies) | 567.67768245 | 68.35768245 |
| 11680 | 998.64 | Scaleway Managed Database for MySQL (one copy) | 730.935765 | -267.704235 |
| 11680 | 998.64 | Scaleway Managed Database for MySQL (two copies) | 1118.33172045 | 119.69172045 |
| 23360 | 1997.28 | Scaleway Managed Database for MySQL (one copy) | 1450.744965 | -546.535035 |
| 23360 | 1997.28 | Scaleway Managed Database for MySQL (two copies) | 2219.63979645 | 222.35979645 |
| 46720 | 3994.56 | Scaleway Managed Database for MySQL (one copy) | 2890.363365 | -1104.196635 |
| 46720 | 3994.56 | Scaleway Managed Database for MySQL (two copies) | 4422.25594845 | 427.69594845 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for MySQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b25b6012b5ed1e8ec2e2a6387ae4012ad5692fbe` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### RDS MySQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/b4406ca4abe7d5aa262d867c4285b230767e42734268c4523ae4009551ef526a-light.svg)

![RDS MySQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/6836050960ac9c073c51fc174656bb19320d30ae7b2af1c5f50cc697d79b426e-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS MySQL | Target scenario | Scaleway Managed Database for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 1.15 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 99.952715 |
| 10 | 1.15 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 153.53715395 |
| 100 | 11.5 | Scaleway Managed Database for MySQL (one copy) | 115.7544975 | 104.2544975 |
| 100 | 11.5 | Scaleway Managed Database for MySQL (two copies) | 177.104381175 | 165.604381175 |
| 1000 | 115 | Scaleway Managed Database for MySQL (one copy) | 1025.26590375 | 910.26590375 |
| 1000 | 115 | Scaleway Managed Database for MySQL (two copies) | 1568.6568327375 | 1453.6568327375 |
| 10000 | 1150 | Scaleway Managed Database for MySQL (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | Scaleway Managed Database for MySQL (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | Scaleway Managed Database for MySQL (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | Scaleway Managed Database for MySQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Scaleway Managed Database for MySQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Scaleway Managed Database for MySQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Scaleway Managed Database for MySQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Scaleway Managed Database for MySQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for MySQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b25b6012b5ed1e8ec2e2a6387ae4012ad5692fbe` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### RDS MySQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/c73d4d232d72ab9b6add02518ba160bb710f628f19625ac2ab0f2f7525b728e9-light.svg)

![RDS MySQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/ed7af1928fc26b5730688bf9a83dd72a09e039456d4f17c008d0f1959d673d05-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS MySQL | Target scenario | Scaleway Managed Database for MySQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 1 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 1 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 10 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 100 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 100 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 1000 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 1000 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |
| 10000 | 0 | Scaleway Managed Database for MySQL (one copy) | 101.102715 | 101.102715 |
| 10000 | 0 | Scaleway Managed Database for MySQL (two copies) | 154.68715395 | 154.68715395 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for MySQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for MySQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for MySQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary compute | 0.0855 | USD/vCPU-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for MySQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for MySQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for MySQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for MySQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for MySQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for MySQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for MySQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for MySQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for MySQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for MySQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for MySQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-b25b6012b5ed1e8ec2e2a6387ae4012ad5692fbe` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

## RDS PostgreSQL → Cloud SQL for PostgreSQL

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `google::1.0.0::cloudsql::postgresql`.

### Monthly database allocation cost · RDS Single-AZ

Reference monthly service charges by Primary vCPUs. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Monthly database allocation cost · RDS Single-AZ: monthly service charges by Primary vCPUs. Exact values are in the following table.](/images/diagrams/service-cost/b2a71eaa296884442683bc38630e3ae7c80a121d92f710f81a4835cbf96adaac-light.svg)

![Monthly database allocation cost · RDS Single-AZ: monthly service charges by Primary vCPUs. Exact values are in the following table.](/images/diagrams/service-cost/d81bfbd53780816ff3f8f41222b134f8c238f80aceb4cb77b5a6bdc26fa2d0a9-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.

| Primary vCPUs (vCPU) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 2 | 141.44 | Cloud SQL for PostgreSQL (zonal) | 118.178 | -23.262 |
| 2 | 141.44 | Cloud SQL for PostgreSQL (HA) | 236.356 | 94.916 |
| 4 | 271.38 | Cloud SQL for PostgreSQL (zonal) | 219.356 | -52.024 |
| 4 | 271.38 | Cloud SQL for PostgreSQL (HA) | 438.712 | 167.332 |
| 8 | 531.26 | Cloud SQL for PostgreSQL (zonal) | 421.712 | -109.548 |
| 8 | 531.26 | Cloud SQL for PostgreSQL (HA) | 843.424 | 312.164 |
| 16 | 1051.02 | Cloud SQL for PostgreSQL (zonal) | 826.424 | -224.596 |
| 16 | 1051.02 | Cloud SQL for PostgreSQL (HA) | 1652.848 | 601.828 |
| 32 | 2090.54 | Cloud SQL for PostgreSQL (zonal) | 1635.848 | -454.692 |
| 32 | 2090.54 | Cloud SQL for PostgreSQL (HA) | 3271.696 | 1181.156 |

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (zonal) | RDS one-standby Multi-AZ | 0 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Provisioned Cloud SQL SSD | 100 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL high availability | 0 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (zonal) | Share of database cost allocated here | 1 | fraction |
| Cloud SQL for PostgreSQL (HA) | RDS one-standby Multi-AZ | 0 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (HA) | Provisioned Cloud SQL SSD | 100 | GiB |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL high availability | 1 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (HA) | Share of database cost allocated here | 1 | fraction |

* RDS PostgreSQL db.m6i family in us-east-1, Single-AZ; Cloud SQL Enterprise PostgreSQL in us-central1. 4 GiB memory per primary vCPU, 100 GiB provisioned storage and 730 running hours on both sides. Equal resource counts are a configuration comparison, not evidence of equal throughput or availability.
* Reference list rates as of 2026-10-18: us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Cloud SQL configuration pricing only: throughput, peak load, latency and storage suitability are unqualified. No capacity benchmark is implied.
* 730 running hours per modeled month. Dedicated database compute and disk cost money even with zero origin traffic.
* Shared database allocation is a caller assumption. Do not sum a full database charge for every adapter sharing that database.
* Source configurations are continuously running PostgreSQL db.m6i instances with 2, 4, 8, 16 or 32 primary vCPUs and 4 GiB RAM per vCPU. Architecture and resource counts do not prove equal performance.
* Application traffic uses native PostgreSQL pgwire directly; no adapter proxy compute is required in that data path. Optional hosting and Tensor9 allocations still require an explicit selection.
* Source primary vCPU-hours and storage exclude the standby; rdsHa adds one copy. Legacy source dollar inputs include the selected standby charge.
* Source usage is priced at reference rates; source lines are not actual invoice charges. Free or discounted usage still consumes target resources.
* RDS PostgreSQL pricing: [https://aws.amazon.com/rds/postgresql/pricing/](https://aws.amazon.com/rds/postgresql/pricing/)
* AWS RDS us-east-1 reference price list: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL Enterprise PostgreSQL pricing: [https://cloud.google.com/sql/pricing/](https://cloud.google.com/sql/pricing/)

Cost scope: Modeled monthly service usage and selected database allocation; adapter hosting and Tensor9 fees only when explicitly supplied.

Excluded costs and workload features:

* Free allowances, negotiated discounts, credits, tax and one-time migration charges.
* Charges not listed in the breakdown, including additional network transfer, backups and encryption/KMS.
* Application compute changes, operational labor and portfolio-wide sharing or deduplication of infrastructure.
* Aurora, other database engines or instance families, RDS Proxy, extra gp3 IOPS/throughput, replicas and Extended Support.

Rate snapshot: 2026-10-18. Model: `rds-postgresql-cloudsql` / `2026-10-18.4`. Curve: `vcpu-rds-ha-0`.

Sources: [RDS PostgreSQL pricing](https://aws.amazon.com/rds/postgresql/pricing/), [AWS RDS us-east-1 reference price list](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL Enterprise PostgreSQL pricing](https://cloud.google.com/sql/pricing/).

### Monthly storage allocation cost · RDS Single-AZ

Reference monthly service charges by Primary provisioned storage. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Monthly storage allocation cost · RDS Single-AZ: monthly service charges by Primary provisioned storage. Exact values are in the following table.](/images/diagrams/service-cost/060d2d90498011920427c4ab82218be119c8393b2eff1d63a72eedee4d1311ce-light.svg)

![Monthly storage allocation cost · RDS Single-AZ: monthly service charges by Primary provisioned storage. Exact values are in the following table.](/images/diagrams/service-cost/92fb2588a8bf4c1071ec2dd210e6f96dbc72f2f15368b2afce31f0f77310d90b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.

| Primary provisioned storage (GiB) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 20 | 132.24 | Cloud SQL for PostgreSQL (zonal) | 104.578 | -27.662 |
| 20 | 132.24 | Cloud SQL for PostgreSQL (HA) | 209.156 | 76.916 |
| 100 | 141.44 | Cloud SQL for PostgreSQL (zonal) | 118.178 | -23.262 |
| 100 | 141.44 | Cloud SQL for PostgreSQL (HA) | 236.356 | 94.916 |
| 1000 | 244.94 | Cloud SQL for PostgreSQL (zonal) | 271.178 | 26.238 |
| 1000 | 244.94 | Cloud SQL for PostgreSQL (HA) | 542.356 | 297.416 |
| 10000 | 1279.94 | Cloud SQL for PostgreSQL (zonal) | 1801.178 | 521.238 |
| 10000 | 1279.94 | Cloud SQL for PostgreSQL (HA) | 3602.356 | 2322.416 |
| 65536 | 7666.58 | Cloud SQL for PostgreSQL (zonal) | 11242.298 | 3575.718 |
| 65536 | 7666.58 | Cloud SQL for PostgreSQL (HA) | 22484.596 | 14818.016 |

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (zonal) | RDS one-standby Multi-AZ | 0 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Provisioned Cloud SQL SSD | 20 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL high availability | 0 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (zonal) | Share of database cost allocated here | 1 | fraction |
| Cloud SQL for PostgreSQL (HA) | RDS one-standby Multi-AZ | 0 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (HA) | Provisioned Cloud SQL SSD | 20 | GiB |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL high availability | 1 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (HA) | Share of database cost allocated here | 1 | fraction |

* Two primary vCPUs and 8 GiB memory, with 730 running hours; RDS Single-AZ versus Cloud SQL zonal and HA. RDS uses baseline gp3 and Cloud SQL uses SSD. IOPS, throughput and backup requirements remain unqualified and are excluded.
* Reference list rates as of 2026-10-18: us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Cloud SQL configuration pricing only: throughput, peak load, latency and storage suitability are unqualified. No capacity benchmark is implied.
* 730 running hours per modeled month. Dedicated database compute and disk cost money even with zero origin traffic.
* Shared database allocation is a caller assumption. Do not sum a full database charge for every adapter sharing that database.
* Source configurations are continuously running PostgreSQL db.m6i instances with 2, 4, 8, 16 or 32 primary vCPUs and 4 GiB RAM per vCPU. Architecture and resource counts do not prove equal performance.
* Application traffic uses native PostgreSQL pgwire directly; no adapter proxy compute is required in that data path. Optional hosting and Tensor9 allocations still require an explicit selection.
* Source primary vCPU-hours and storage exclude the standby; rdsHa adds one copy. Legacy source dollar inputs include the selected standby charge.
* Source usage is priced at reference rates; source lines are not actual invoice charges. Free or discounted usage still consumes target resources.
* RDS PostgreSQL pricing: [https://aws.amazon.com/rds/postgresql/pricing/](https://aws.amazon.com/rds/postgresql/pricing/)
* AWS RDS us-east-1 reference price list: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL Enterprise PostgreSQL pricing: [https://cloud.google.com/sql/pricing/](https://cloud.google.com/sql/pricing/)

Cost scope: Modeled monthly service usage and selected database allocation; adapter hosting and Tensor9 fees only when explicitly supplied.

Excluded costs and workload features:

* Free allowances, negotiated discounts, credits, tax and one-time migration charges.
* Charges not listed in the breakdown, including additional network transfer, backups and encryption/KMS.
* Application compute changes, operational labor and portfolio-wide sharing or deduplication of infrastructure.
* Aurora, other database engines or instance families, RDS Proxy, extra gp3 IOPS/throughput, replicas and Extended Support.

Rate snapshot: 2026-10-18. Model: `rds-postgresql-cloudsql` / `2026-10-18.4`. Curve: `storage-rds-ha-0`.

Sources: [RDS PostgreSQL pricing](https://aws.amazon.com/rds/postgresql/pricing/), [AWS RDS us-east-1 reference price list](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL Enterprise PostgreSQL pricing](https://cloud.google.com/sql/pricing/).

### Monthly database allocation cost · RDS Multi-AZ, one standby

Reference monthly service charges by Primary vCPUs. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Monthly database allocation cost · RDS Multi-AZ, one standby: monthly service charges by Primary vCPUs. Exact values are in the following table.](/images/diagrams/service-cost/36a8e060e004f622b5e1556057ab8dd741ba41e4aecae6b47584154b64f803a4-light.svg)

![Monthly database allocation cost · RDS Multi-AZ, one standby: monthly service charges by Primary vCPUs. Exact values are in the following table.](/images/diagrams/service-cost/6ad54dc48852b7aee8a1015d8c853ec7b3a40a50ed43dadb51645f756f19cfcb-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.

| Primary vCPUs (vCPU) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 2 | 282.88 | Cloud SQL for PostgreSQL (zonal) | 118.178 | -164.702 |
| 2 | 282.88 | Cloud SQL for PostgreSQL (HA) | 236.356 | -46.524 |
| 4 | 542.76 | Cloud SQL for PostgreSQL (zonal) | 219.356 | -323.404 |
| 4 | 542.76 | Cloud SQL for PostgreSQL (HA) | 438.712 | -104.048 |
| 8 | 1062.52 | Cloud SQL for PostgreSQL (zonal) | 421.712 | -640.808 |
| 8 | 1062.52 | Cloud SQL for PostgreSQL (HA) | 843.424 | -219.096 |
| 16 | 2102.04 | Cloud SQL for PostgreSQL (zonal) | 826.424 | -1275.616 |
| 16 | 2102.04 | Cloud SQL for PostgreSQL (HA) | 1652.848 | -449.192 |
| 32 | 4181.08 | Cloud SQL for PostgreSQL (zonal) | 1635.848 | -2545.232 |
| 32 | 4181.08 | Cloud SQL for PostgreSQL (HA) | 3271.696 | -909.384 |

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (zonal) | RDS one-standby Multi-AZ | 1 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Provisioned Cloud SQL SSD | 100 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL high availability | 0 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (zonal) | Share of database cost allocated here | 1 | fraction |
| Cloud SQL for PostgreSQL (HA) | RDS one-standby Multi-AZ | 1 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (HA) | Provisioned Cloud SQL SSD | 100 | GiB |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL high availability | 1 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (HA) | Share of database cost allocated here | 1 | fraction |

* RDS PostgreSQL db.m6i family in us-east-1, Multi-AZ, one standby; Cloud SQL Enterprise PostgreSQL in us-central1. 4 GiB memory per primary vCPU, 100 GiB provisioned storage and 730 running hours on both sides. Equal resource counts are a configuration comparison, not evidence of equal throughput or availability.
* Reference list rates as of 2026-10-18: us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Cloud SQL configuration pricing only: throughput, peak load, latency and storage suitability are unqualified. No capacity benchmark is implied.
* 730 running hours per modeled month. Dedicated database compute and disk cost money even with zero origin traffic.
* Shared database allocation is a caller assumption. Do not sum a full database charge for every adapter sharing that database.
* Source configurations are continuously running PostgreSQL db.m6i instances with 2, 4, 8, 16 or 32 primary vCPUs and 4 GiB RAM per vCPU. Architecture and resource counts do not prove equal performance.
* Application traffic uses native PostgreSQL pgwire directly; no adapter proxy compute is required in that data path. Optional hosting and Tensor9 allocations still require an explicit selection.
* Source primary vCPU-hours and storage exclude the standby; rdsHa adds one copy. Legacy source dollar inputs include the selected standby charge.
* Source usage is priced at reference rates; source lines are not actual invoice charges. Free or discounted usage still consumes target resources.
* RDS PostgreSQL pricing: [https://aws.amazon.com/rds/postgresql/pricing/](https://aws.amazon.com/rds/postgresql/pricing/)
* AWS RDS us-east-1 reference price list: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL Enterprise PostgreSQL pricing: [https://cloud.google.com/sql/pricing/](https://cloud.google.com/sql/pricing/)

Cost scope: Modeled monthly service usage and selected database allocation; adapter hosting and Tensor9 fees only when explicitly supplied.

Excluded costs and workload features:

* Free allowances, negotiated discounts, credits, tax and one-time migration charges.
* Charges not listed in the breakdown, including additional network transfer, backups and encryption/KMS.
* Application compute changes, operational labor and portfolio-wide sharing or deduplication of infrastructure.
* Aurora, other database engines or instance families, RDS Proxy, extra gp3 IOPS/throughput, replicas and Extended Support.

Rate snapshot: 2026-10-18. Model: `rds-postgresql-cloudsql` / `2026-10-18.4`. Curve: `vcpu-rds-ha-1`.

Sources: [RDS PostgreSQL pricing](https://aws.amazon.com/rds/postgresql/pricing/), [AWS RDS us-east-1 reference price list](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL Enterprise PostgreSQL pricing](https://cloud.google.com/sql/pricing/).

### Monthly storage allocation cost · RDS Multi-AZ, one standby

Reference monthly service charges by Primary provisioned storage. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![Monthly storage allocation cost · RDS Multi-AZ, one standby: monthly service charges by Primary provisioned storage. Exact values are in the following table.](/images/diagrams/service-cost/dd1d8fabedb3c0adb084ceef76b48be7484395ff2953af39c15ada8bafb8de65-light.svg)

![Monthly storage allocation cost · RDS Multi-AZ, one standby: monthly service charges by Primary provisioned storage. Exact values are in the following table.](/images/diagrams/service-cost/c1b0906bd9c25012d1613feef92f02895b5dfbf80576bda8ef9a9b27470127ea-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.

| Primary provisioned storage (GiB) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 20 | 264.48 | Cloud SQL for PostgreSQL (zonal) | 104.578 | -159.902 |
| 20 | 264.48 | Cloud SQL for PostgreSQL (HA) | 209.156 | -55.324 |
| 100 | 282.88 | Cloud SQL for PostgreSQL (zonal) | 118.178 | -164.702 |
| 100 | 282.88 | Cloud SQL for PostgreSQL (HA) | 236.356 | -46.524 |
| 1000 | 489.88 | Cloud SQL for PostgreSQL (zonal) | 271.178 | -218.702 |
| 1000 | 489.88 | Cloud SQL for PostgreSQL (HA) | 542.356 | 52.476 |
| 10000 | 2559.88 | Cloud SQL for PostgreSQL (zonal) | 1801.178 | -758.702 |
| 10000 | 2559.88 | Cloud SQL for PostgreSQL (HA) | 3602.356 | 1042.476 |
| 65536 | 15333.16 | Cloud SQL for PostgreSQL (zonal) | 11242.298 | -4090.862 |
| 65536 | 15333.16 | Cloud SQL for PostgreSQL (HA) | 22484.596 | 7151.436 |

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (zonal) | RDS one-standby Multi-AZ | 1 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Provisioned Cloud SQL SSD | 20 | GiB |
| Cloud SQL for PostgreSQL (zonal) | Cloud SQL high availability | 0 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (zonal) | Share of database cost allocated here | 1 | fraction |
| Cloud SQL for PostgreSQL (HA) | RDS one-standby Multi-AZ | 1 | 0 = Single-AZ, 1 = Multi-AZ |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL vCPUs | 2 | vCPU |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL memory | 8 | GiB |
| Cloud SQL for PostgreSQL (HA) | Provisioned Cloud SQL SSD | 20 | GiB |
| Cloud SQL for PostgreSQL (HA) | Cloud SQL high availability | 1 | 0 = zonal, 1 = HA |
| Cloud SQL for PostgreSQL (HA) | Share of database cost allocated here | 1 | fraction |

* Two primary vCPUs and 8 GiB memory, with 730 running hours; RDS Multi-AZ, one standby versus Cloud SQL zonal and HA. RDS uses baseline gp3 and Cloud SQL uses SSD. IOPS, throughput and backup requirements remain unqualified and are excluded.
* Reference list rates as of 2026-10-18: us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Cloud SQL configuration pricing only: throughput, peak load, latency and storage suitability are unqualified. No capacity benchmark is implied.
* 730 running hours per modeled month. Dedicated database compute and disk cost money even with zero origin traffic.
* Shared database allocation is a caller assumption. Do not sum a full database charge for every adapter sharing that database.
* Source configurations are continuously running PostgreSQL db.m6i instances with 2, 4, 8, 16 or 32 primary vCPUs and 4 GiB RAM per vCPU. Architecture and resource counts do not prove equal performance.
* Application traffic uses native PostgreSQL pgwire directly; no adapter proxy compute is required in that data path. Optional hosting and Tensor9 allocations still require an explicit selection.
* Source primary vCPU-hours and storage exclude the standby; rdsHa adds one copy. Legacy source dollar inputs include the selected standby charge.
* Source usage is priced at reference rates; source lines are not actual invoice charges. Free or discounted usage still consumes target resources.
* RDS PostgreSQL pricing: [https://aws.amazon.com/rds/postgresql/pricing/](https://aws.amazon.com/rds/postgresql/pricing/)
* AWS RDS us-east-1 reference price list: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Cloud SQL Enterprise PostgreSQL pricing: [https://cloud.google.com/sql/pricing/](https://cloud.google.com/sql/pricing/)

Cost scope: Modeled monthly service usage and selected database allocation; adapter hosting and Tensor9 fees only when explicitly supplied.

Excluded costs and workload features:

* Free allowances, negotiated discounts, credits, tax and one-time migration charges.
* Charges not listed in the breakdown, including additional network transfer, backups and encryption/KMS.
* Application compute changes, operational labor and portfolio-wide sharing or deduplication of infrastructure.
* Aurora, other database engines or instance families, RDS Proxy, extra gp3 IOPS/throughput, replicas and Extended Support.

Rate snapshot: 2026-10-18. Model: `rds-postgresql-cloudsql` / `2026-10-18.4`. Curve: `storage-rds-ha-1`.

Sources: [RDS PostgreSQL pricing](https://aws.amazon.com/rds/postgresql/pricing/), [AWS RDS us-east-1 reference price list](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Cloud SQL Enterprise PostgreSQL pricing](https://cloud.google.com/sql/pricing/).

## RDS PostgreSQL → Cloud SQL for PostgreSQL (RDS proxy)

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `google::1.0.0::cloudsql::postgresql::proxied`.

### RDS PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/062fa49b95e51345bb408c8ac33e80487a0098fb7c437ebf4c6ac7509293c0c9-light.svg)

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/d606e8ac881d2f94fffe9bd753ff0a3b05f4d9ff2142a40473fe9451e9ff5cef-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1460 | 129.94 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | -4.9 |
| 1460 | 129.94 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 120.14 |
| 2920 | 259.88 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 233.08 | -26.8 |
| 2920 | 259.88 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 466.16 | 206.28 |
| 5840 | 519.76 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 449.16 | -70.6 |
| 5840 | 519.76 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 898.32 | 378.56 |
| 11680 | 1039.52 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 881.32 | -158.2 |
| 11680 | 1039.52 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 1762.64 | 723.12 |
| 23360 | 2079.04 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 1745.64 | -333.4 |
| 23360 | 2079.04 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 3491.28 | 1412.24 |
| 46720 | 4158.08 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 3474.28 | -683.8 |
| 46720 | 4158.08 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 6948.56 | 2790.48 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::cloudsql::postgresql::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-1dbf7e73343e44effdc50f2b5f5a2420ba0465c3` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

### RDS PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/346bcb661f759ce51500489fdddae05cf38aa6010a10b7c938fd992db711fa19-light.svg)

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/ca90c2234fd24c8a60647940826f02eafad6e02b89a5ac2a0e8c76cdc94cd455-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10 | 1.15 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 123.89 |
| 10 | 1.15 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 248.93 |
| 100 | 11.5 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 142.3 | 130.8 |
| 100 | 11.5 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 284.6 | 273.1 |
| 1000 | 115 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 1238.75 | 1123.75 |
| 1000 | 115 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 2477.5 | 2362.5 |
| 10000 | 1150 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::cloudsql::postgresql::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-1dbf7e73343e44effdc50f2b5f5a2420ba0465c3` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

### RDS PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: Explicit target planning region; currency: USD.

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/f611f8479905f3be5911d406c7f83173a79913174b207b827e2d3da708776853-light.svg)

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/766a2220198e529ebc3873f27f7417a8a640991ecc0b87c44abe51af673444c2-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS PostgreSQL | Target scenario | Cloud SQL for PostgreSQL (RDS proxy) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 0 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 1 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 10 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 100 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 100 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 1000 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 1000 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |
| 10000 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (one copy) | 125.04 | 125.04 |
| 10000 | 0 | Cloud SQL for PostgreSQL (RDS proxy) (two copies) | 250.08 | 250.08 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for PostgreSQL (RDS proxy) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Cloud SQL for PostgreSQL (RDS proxy) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target CPU / source CPU | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target RAM / source RAM | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target disk / source disk | 1 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · read | 0 | USD/read |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · write | 0 | USD/write |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-month | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · network GiB | 0.09 | USD/network GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · vCPU-hour | 0.05 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · GiB-hour | 0.006 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · input token | 0 | USD/input token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target price guess · output token | 0 | USD/output token |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Target monthly base charge | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Approximate target service minimum | 0 | USD/month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary CPU | 2 | vCPU |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary RAM | 8 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Minimum primary disk | 100 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning peak / mean work | 2 | multiplier |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Planning disk utilization | 0.8 | fraction |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Cloud SQL for PostgreSQL (RDS proxy) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → Explicit target planning region, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* All target rates are editable authored planning guesses, not verified current list prices, customer discounts or benchmark evidence.
* Unknown target google::1.0.0::cloudsql::postgresql::proxied: conservative generic provisioned planning fallback. Target-specific tariffs, resource constraints and deployment behavior are unqualified.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-1dbf7e73343e44effdc50f2b5f5a2420ba0465c3` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json).

## RDS PostgreSQL → OCI Database with PostgreSQL

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `oci::1.0.0::psql`.

### RDS PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/6b311f93c324a30bd86598ec2073adcb864932ef1451c831637c066346e811dd-light.svg)

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/857c6f66bef2f1db4c00ca7d0fa12a71237b3cc5d19d0be33c3a55d23a92b9a9-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1460 | 129.94 | OCI Database with PostgreSQL (one copy) | 94.9 | -35.04 |
| 1460 | 129.94 | OCI Database with PostgreSQL (two copies) | 173.64 | 43.7 |
| 2920 | 259.88 | OCI Database with PostgreSQL (one copy) | 173.64 | -86.24 |
| 2920 | 259.88 | OCI Database with PostgreSQL (two copies) | 340.08 | 80.2 |
| 5840 | 519.76 | OCI Database with PostgreSQL (one copy) | 340.08 | -179.68 |
| 5840 | 519.76 | OCI Database with PostgreSQL (two copies) | 672.96 | 153.2 |
| 11680 | 1039.52 | OCI Database with PostgreSQL (one copy) | 672.96 | -366.56 |
| 11680 | 1039.52 | OCI Database with PostgreSQL (two copies) | 1338.72 | 299.2 |
| 23360 | 2079.04 | OCI Database with PostgreSQL (one copy) | 1338.72 | -740.32 |
| 23360 | 2079.04 | OCI Database with PostgreSQL (two copies) | 2670.24 | 591.2 |
| 46720 | 4158.08 | OCI Database with PostgreSQL (one copy) | 2670.24 | -1487.84 |
| 46720 | 4158.08 | OCI Database with PostgreSQL (two copies) | 5333.28 | 1175.2 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d81a7cca1998e220a08619ab9bb758089ead10f3` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

### RDS PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/6a2f095abd82113cc919511ea09583f1b5200ebaa6ad8975fa83a93aba36302c-light.svg)

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/a69408792f02fdadc9b26cf42075a19aa641829c74a454d56a924228587d4d4b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10 | 1.15 | OCI Database with PostgreSQL (one copy) | 94.9 | 93.75 |
| 10 | 1.15 | OCI Database with PostgreSQL (two copies) | 173.64 | 172.49 |
| 100 | 11.5 | OCI Database with PostgreSQL (one copy) | 96.94 | 85.44 |
| 100 | 11.5 | OCI Database with PostgreSQL (two copies) | 183.08 | 171.58 |
| 1000 | 115 | OCI Database with PostgreSQL (one copy) | 810.12 | 695.12 |
| 1000 | 115 | OCI Database with PostgreSQL (two copies) | 1528.44 | 1413.44 |
| 10000 | 1150 | OCI Database with PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | OCI Database with PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | OCI Database with PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | OCI Database with PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / OCI Database with PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / OCI Database with PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / OCI Database with PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / OCI Database with PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d81a7cca1998e220a08619ab9bb758089ead10f3` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

### RDS PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-ashburn-1 (global list); currency: USD.

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/41bcc5106ac8a8ce0bbd26c7574856fa7e28de101955387fb1a3abfa93b91beb-light.svg)

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/b03b6fcc3259841e5c1c162d13659bf0151c0a4645a5c72cae1315de851e6f61-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS PostgreSQL | Target scenario | OCI Database with PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 0 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 1 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 10 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 100 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 100 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 1000 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 1000 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |
| 10000 | 0 | OCI Database with PostgreSQL (one copy) | 94.9 | 94.9 |
| 10000 | 0 | OCI Database with PostgreSQL (two copies) | 173.64 | 173.64 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI Database with PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / OCI Database with PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (one copy) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (one copy) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (one copy) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| OCI Database with PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| OCI Database with PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| OCI Database with PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| OCI Database with PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| OCI Database with PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| OCI Database with PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-month | 0.072 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| OCI Database with PostgreSQL (two copies) | Target list price · network GiB | 0 | USD/network GiB |
| OCI Database with PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| OCI Database with PostgreSQL (two copies) | Target list price · vCPU-hour | 0.049 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Target list price · GiB-hour | 0.002 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| OCI Database with PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| OCI Database with PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| OCI Database with PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| OCI Database with PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| OCI Database with PostgreSQL (two copies) | Target service minimum | 94.9 | USD/month |
| OCI Database with PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| OCI Database with PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| OCI Database with PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| OCI Database with PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| OCI Database with PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| OCI Database with PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| OCI Database with PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| OCI Database with PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| OCI Database with PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| OCI Database with PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| OCI Database with PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| OCI Database with PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| OCI Database with PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| OCI Database with PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| OCI Database with PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-ashburn-1 (global list), USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* OCI Database with PostgreSQL bills provisioned OCPUs (one x86 OCPU is two vCPUs), a memory fee and Database Optimized Storage by used GB. The storage is one pool replicated across availability domains and shared by every node, so it is billed once however many copies run. The memory rate is an assumption (see its source). The smallest shape is PostgreSQL.VM.Standard.E5.Flex with 1 OCPU and 16 GB ([https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm](https://docs.oracle.com/en-us/iaas/Content/postgresql/supported-shapes.htm)). Backups bill as Object Storage and are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* OCI price list API: [https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/)
* Oracle Cloud price list: [https://www.oracle.com/cloud/price-list/](https://www.oracle.com/cloud/price-list/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d81a7cca1998e220a08619ab9bb758089ead10f3` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [OCI price list API](https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/), [Oracle Cloud price list](https://www.oracle.com/cloud/price-list/).

## RDS PostgreSQL → PostgreSQL (CloudNativePG)

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `cloudnative::1.0.0::postgresql`.

### RDS PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/fa9fd6f1b3554ee1ccda517d425df7465ec3ad3ad86bfcff8d7446634f703478-light.svg)

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/908ce1924549291d49171e12dbffcddbab7157a8ac6d12c6d14be78844068614-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1460 | 129.94 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | -49.04386 |
| 1460 | 129.94 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 31.85228 |
| 2920 | 259.88 | PostgreSQL (CloudNativePG) (one copy) | 151.79228 | -108.08772 |
| 2920 | 259.88 | PostgreSQL (CloudNativePG) (two copies) | 303.58456 | 43.70456 |
| 5840 | 519.76 | PostgreSQL (CloudNativePG) (one copy) | 293.58456 | -226.17544 |
| 5840 | 519.76 | PostgreSQL (CloudNativePG) (two copies) | 587.16912 | 67.40912 |
| 11680 | 1039.52 | PostgreSQL (CloudNativePG) (one copy) | 577.16912 | -462.35088 |
| 11680 | 1039.52 | PostgreSQL (CloudNativePG) (two copies) | 1154.33824 | 114.81824 |
| 23360 | 2079.04 | PostgreSQL (CloudNativePG) (one copy) | 1144.33824 | -934.70176 |
| 23360 | 2079.04 | PostgreSQL (CloudNativePG) (two copies) | 2288.67648 | 209.63648 |
| 46720 | 4158.08 | PostgreSQL (CloudNativePG) (one copy) | 2278.67648 | -1879.40352 |
| 46720 | 4158.08 | PostgreSQL (CloudNativePG) (two copies) | 4557.35296 | 399.27296 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-e86e5324ae76bdb19c4a0e6a1eed4450d01c44e5` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### RDS PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/b344fb487da025dd629e17b1405826a72aa1a112e9219610fc4cf7b73c5f5335-light.svg)

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/37050f310bc5a62e457d448fe3c447acca972cfb89ec4c04d31c50a06c6aba43-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10 | 1.15 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 79.74614 |
| 10 | 1.15 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 160.64228 |
| 100 | 11.5 | PostgreSQL (CloudNativePG) (one copy) | 92.08216 | 80.58216 |
| 100 | 11.5 | PostgreSQL (CloudNativePG) (two copies) | 184.16432 | 172.66432 |
| 1000 | 115 | PostgreSQL (CloudNativePG) (one copy) | 806.01748 | 691.01748 |
| 1000 | 115 | PostgreSQL (CloudNativePG) (two copies) | 1612.03496 | 1497.03496 |
| 10000 | 1150 | PostgreSQL (CloudNativePG) (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | PostgreSQL (CloudNativePG) (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | PostgreSQL (CloudNativePG) (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | PostgreSQL (CloudNativePG) (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / PostgreSQL (CloudNativePG) (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / PostgreSQL (CloudNativePG) (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-e86e5324ae76bdb19c4a0e6a1eed4450d01c44e5` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

### RDS PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: us-central1; currency: USD.

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/b486662e51369d39b65c09a8e78c15f23b5191236346cc9232947cc29d773415-light.svg)

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/630a4f32a1bd23fd494de9b0d8cc454e8907935d94a5184d9f5fcd2aa0a817ef-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS PostgreSQL | Target scenario | PostgreSQL (CloudNativePG) | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 0 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 1 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 10 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 100 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 100 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 1000 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 1000 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |
| 10000 | 0 | PostgreSQL (CloudNativePG) (one copy) | 80.89614 | 80.89614 |
| 10000 | 0 | PostgreSQL (CloudNativePG) (two copies) | 161.79228 | 161.79228 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL (CloudNativePG) (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL (CloudNativePG) (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL (CloudNativePG) (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL (CloudNativePG) (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-month | 0.1 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · network GiB | 0.12 | USD/network GiB |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · vCPU-hour | 0.031611 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target list price · GiB-hour | 0.004237 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL (CloudNativePG) (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL (CloudNativePG) (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Approximate target service minimum | 0 | USD/month |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL (CloudNativePG) (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL (CloudNativePG) (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL (CloudNativePG) (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL (CloudNativePG) (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL (CloudNativePG) (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL (CloudNativePG) (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL (CloudNativePG) (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → us-central1, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Self-hosted target: CloudNativePG (Apache-2.0, [https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE](https://github.com/cloudnative-pg/cloudnative-pg/blob/main/LICENSE)) with PostgreSQL. Selected CPU, RAM and physical disk allocation grow with workload; no per-operation price is assumed. No vendor tariff: the target runs on the customer's Kubernetes nodes, so CPU, RAM and disk bill at the Kubernetes node reference rates, Compute Engine N2 in us-central1 (USD 0.031611/vCPU-hour, 0.004237/GiB-hour) with pd-balanced disk (0.10/GiB-month) and Premium Tier egress. The cost-optimized E2 series lists 0.02181159/vCPU-hour and 0.00292353/GiB-hour; AWS m6i/r6i and Azure D2s v5/E2s v5 both decompose to 0.033/vCPU-hour and 0.00375/GiB-hour. Requested CPU and RAM are billed; node bin-packing waste, system reservations and the cluster control-plane fee are not.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Compute Engine general-purpose machine pricing: [https://cloud.google.com/products/compute/pricing/general-purpose](https://cloud.google.com/products/compute/pricing/general-purpose)
* Disk and image pricing: [https://cloud.google.com/compute/disks-image-pricing](https://cloud.google.com/compute/disks-image-pricing)
* Network pricing: [https://cloud.google.com/vpc/network-pricing](https://cloud.google.com/vpc/network-pricing)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-e86e5324ae76bdb19c4a0e6a1eed4450d01c44e5` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Compute Engine general-purpose machine pricing](https://cloud.google.com/products/compute/pricing/general-purpose), [Disk and image pricing](https://cloud.google.com/compute/disks-image-pricing), [Network pricing](https://cloud.google.com/vpc/network-pricing).

## RDS PostgreSQL → PostgreSQL Flexible Server

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `azure::1.0.0::postgresql::flexible`.

### RDS PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/c4c53d7d93e34a87be5bffb493747d94fefc5cb50553c8f7ec5cf2426d876fdf-light.svg)

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/33110e2b9c3aa516d68ded315b39e2124aa5c84491a9008174b023439cb314d8-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1460 | 129.94 | PostgreSQL Flexible Server (one copy) | 141.44 | 11.5 |
| 1460 | 129.94 | PostgreSQL Flexible Server (two copies) | 282.88 | 152.94 |
| 2920 | 259.88 | PostgreSQL Flexible Server (one copy) | 271.38 | 11.5 |
| 2920 | 259.88 | PostgreSQL Flexible Server (two copies) | 542.76 | 282.88 |
| 5840 | 519.76 | PostgreSQL Flexible Server (one copy) | 531.26 | 11.5 |
| 5840 | 519.76 | PostgreSQL Flexible Server (two copies) | 1062.52 | 542.76 |
| 11680 | 1039.52 | PostgreSQL Flexible Server (one copy) | 1051.02 | 11.5 |
| 11680 | 1039.52 | PostgreSQL Flexible Server (two copies) | 2102.04 | 1062.52 |
| 23360 | 2079.04 | PostgreSQL Flexible Server (one copy) | 2090.54 | 11.5 |
| 23360 | 2079.04 | PostgreSQL Flexible Server (two copies) | 4181.08 | 2102.04 |
| 46720 | 4158.08 | PostgreSQL Flexible Server (one copy) | 4169.58 | 11.5 |
| 46720 | 4158.08 | PostgreSQL Flexible Server (two copies) | 8339.16 | 4181.08 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d1ce786dca96f39d19183d1841f6720bd4461047` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### RDS PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/d23b2d2e493f2b2e4521576b088a54af51b26f4d444c18d0e78dbc3d76760ed3-light.svg)

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/80a4f1599156a0c76c23fcc97a31dc3ae0b0a7cd85a4070b3c235ca99a2a1a54-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10 | 1.15 | PostgreSQL Flexible Server (one copy) | 141.44 | 140.29 |
| 10 | 1.15 | PostgreSQL Flexible Server (two copies) | 282.88 | 281.73 |
| 100 | 11.5 | PostgreSQL Flexible Server (one copy) | 160.33 | 148.83 |
| 100 | 11.5 | PostgreSQL Flexible Server (two copies) | 320.66 | 309.16 |
| 1000 | 115 | PostgreSQL Flexible Server (one copy) | 1422.665 | 1307.665 |
| 1000 | 115 | PostgreSQL Flexible Server (two copies) | 2845.33 | 2730.33 |
| 10000 | 1150 | PostgreSQL Flexible Server (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | PostgreSQL Flexible Server (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | PostgreSQL Flexible Server (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | PostgreSQL Flexible Server (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / PostgreSQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / PostgreSQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / PostgreSQL Flexible Server (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / PostgreSQL Flexible Server (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d1ce786dca96f39d19183d1841f6720bd4461047` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

### RDS PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: eastus; currency: USD.

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/109468ac9d0ef4468f7985b46ad64c004c5129b46d2b180861acf0fd8eab7f8c-light.svg)

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/951d2e476b4fadb135b07e9d8983ef2db93ce8529ed25e9407076b331360e63a-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS PostgreSQL | Target scenario | PostgreSQL Flexible Server | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 0 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 1 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 10 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 100 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 100 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 1000 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 1000 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |
| 10000 | 0 | PostgreSQL Flexible Server (one copy) | 141.44 | 141.44 |
| 10000 | 0 | PostgreSQL Flexible Server (two copies) | 282.88 | 282.88 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL Flexible Server (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / PostgreSQL Flexible Server (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (one copy) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (one copy) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (one copy) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (one copy) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (one copy) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (one copy) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (one copy) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (one copy) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (one copy) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (one copy) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (one copy) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (one copy) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| PostgreSQL Flexible Server (one copy) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (one copy) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (one copy) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (one copy) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| PostgreSQL Flexible Server (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database reads | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Origin list price · Database writes | 0 | USD/operations |
| PostgreSQL Flexible Server (two copies) | Target CPU / source CPU | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target RAM / source RAM | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target disk / source disk | 1 | multiplier |
| PostgreSQL Flexible Server (two copies) | Target price guess · read | 0 | USD/read |
| PostgreSQL Flexible Server (two copies) | Target price guess · write | 0 | USD/write |
| PostgreSQL Flexible Server (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-month | 0.115 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| PostgreSQL Flexible Server (two copies) | Target list price · network GiB | 0.087 | USD/network GiB |
| PostgreSQL Flexible Server (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| PostgreSQL Flexible Server (two copies) | Target list price · vCPU-hour | 0.053 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Target list price · GiB-hour | 0.009 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| PostgreSQL Flexible Server (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| PostgreSQL Flexible Server (two copies) | Target price guess · input token | 0 | USD/input token |
| PostgreSQL Flexible Server (two copies) | Target price guess · output token | 0 | USD/output token |
| PostgreSQL Flexible Server (two copies) | Target monthly base charge | 0 | USD/month |
| PostgreSQL Flexible Server (two copies) | Target service minimum | 16.09 | USD/month |
| PostgreSQL Flexible Server (two copies) | Minimum primary CPU | 2 | vCPU |
| PostgreSQL Flexible Server (two copies) | Minimum primary RAM | 8 | GiB |
| PostgreSQL Flexible Server (two copies) | Minimum primary disk | 100 | GiB |
| PostgreSQL Flexible Server (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| PostgreSQL Flexible Server (two copies) | Allocated share of provisioned cost | 1 | fraction |
| PostgreSQL Flexible Server (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| PostgreSQL Flexible Server (two copies) | Planning peak / mean work | 2 | multiplier |
| PostgreSQL Flexible Server (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| PostgreSQL Flexible Server (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| PostgreSQL Flexible Server (two copies) | Planning disk utilization | 0.8 | fraction |
| PostgreSQL Flexible Server (two copies) | Maintenance and log disk reserve | 20 | GiB |
| PostgreSQL Flexible Server (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| PostgreSQL Flexible Server (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| PostgreSQL Flexible Server (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| PostgreSQL Flexible Server (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → eastus, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Azure Database for PostgreSQL Flexible Server bills vCores with a fixed memory ratio per series; the vCPU and memory rates decompose the General Purpose and Memory Optimized v5 prices. High availability bills the standby's compute and storage as a second copy. The minimum is the smallest Burstable server, which does not support high availability; the smallest General Purpose server (2 vCores) costs USD 133.62 per month. Additional IOPS, backup storage and reserved capacity are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Azure Retail Prices API: [https://prices.azure.com/api/retail/prices](https://prices.azure.com/api/retail/prices)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-d1ce786dca96f39d19183d1841f6720bd4461047` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Azure Retail Prices API](https://prices.azure.com/api/retail/prices).

## RDS PostgreSQL → Scaleway Managed Database for PostgreSQL

Origin API: `aws::1.0.0::rds::postgresql`. Target backend: `scaleway::1.0.0::rdb::postgresql`.

**Native or deployment mapping.** This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA. Cost estimates describe the selected service resources; operation coverage remains separate.

### RDS PostgreSQL · Primary allocated compute

Estimated monthly service charges by Primary allocated compute. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/057b8629dfb4ccea2ad15fc431626319de2f97815ae23016c53b285d17fbc72c-light.svg)

![RDS PostgreSQL · Primary allocated compute: monthly service charges by Primary allocated compute. Exact values are in the following table.](/images/diagrams/service-cost/2079b39be395f752d3da4fea1b1458cfd7a6bbf6953c3661a21b72a0a57f48a3-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Primary allocated compute (vCPU-hours/month) | RDS PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1460 | 129.94 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | -28.837285 |
| 1460 | 129.94 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 24.74715395 |
| 2920 | 259.88 | Scaleway Managed Database for PostgreSQL (one copy) | 191.078865 | -68.801135 |
| 2920 | 259.88 | Scaleway Managed Database for PostgreSQL (two copies) | 292.35066345 | 32.47066345 |
| 5840 | 519.76 | Scaleway Managed Database for PostgreSQL (one copy) | 371.031165 | -148.728835 |
| 5840 | 519.76 | Scaleway Managed Database for PostgreSQL (two copies) | 567.67768245 | 47.91768245 |
| 11680 | 1039.52 | Scaleway Managed Database for PostgreSQL (one copy) | 730.935765 | -308.584235 |
| 11680 | 1039.52 | Scaleway Managed Database for PostgreSQL (two copies) | 1118.33172045 | 78.81172045 |
| 23360 | 2079.04 | Scaleway Managed Database for PostgreSQL (one copy) | 1450.744965 | -628.295035 |
| 23360 | 2079.04 | Scaleway Managed Database for PostgreSQL (two copies) | 2219.63979645 | 140.59979645 |
| 46720 | 4158.08 | Scaleway Managed Database for PostgreSQL (one copy) | 2890.363365 | -1267.716635 |
| 46720 | 4158.08 | Scaleway Managed Database for PostgreSQL (two copies) | 4422.25594845 | 264.17594845 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1460 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 2 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 2920 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 4 | 4 | 16 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 5840 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 8 | 8 | 32 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 11680 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 16 | 16 | 64 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 23360 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 32 | 32 | 128 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 46720 vCPU-hours/month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 64 | 64 | 256 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* CPU allocation does not imply an achieved transaction rate. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-c828ea2296c34e1fe5b1f54217e1ed4e0bda56ff` / `2026-10-22.planning.2`. Curve: `workload-1`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### RDS PostgreSQL · Stored data

Estimated monthly service charges by Stored data. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/7901afb4ca60d598da7c42656ed559a5280ef362a7b7f5609c4e4dcdcc7dc02a-light.svg)

![RDS PostgreSQL · Stored data: monthly service charges by Stored data. Exact values are in the following table.](/images/diagrams/service-cost/0cb0dcc6e5a227e57d84b724514fde14a09fdcb8cce47eebc044c6667791dd2b-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Stored data (GiB-month) | RDS PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 1.15 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 99.952715 |
| 10 | 1.15 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 153.53715395 |
| 100 | 11.5 | Scaleway Managed Database for PostgreSQL (one copy) | 115.7544975 | 104.2544975 |
| 100 | 11.5 | Scaleway Managed Database for PostgreSQL (two copies) | 177.104381175 | 165.604381175 |
| 1000 | 115 | Scaleway Managed Database for PostgreSQL (one copy) | 1025.26590375 | 910.26590375 |
| 1000 | 115 | Scaleway Managed Database for PostgreSQL (two copies) | 1568.6568327375 | 1453.6568327375 |
| 10000 | 1150 | Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | Unavailable |
| 10000 | 1150 | Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | Unavailable |
| 100000 | 11500 | Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | Unavailable |
| 100000 | 11500 | Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | Unavailable |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 10 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 10 | 150 | 100 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 16 | 100 | 1275 | 1000 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 10000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 10000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM). |
| 100000 GiB-month / Scaleway Managed Database for PostgreSQL (one copy) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |
| 100000 GiB-month / Scaleway Managed Database for PostgreSQL (two copies) | Unpriced | 0 / 0 | 0 | Unpriced | Unpriced | Unpriced | 100000 | `service-adapter-planning-2026-10-22.1` | Demand exceeds the planning model's joint CPU/RAM allocation bounds (96 vCPU, 624 GiB RAM).; Required disk exceeds the planning model's 65,536 GiB bound. |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Request meters remain at their independently authored defaults. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-c828ea2296c34e1fe5b1f54217e1ed4e0bda56ff` / `2026-10-22.planning.2`. Curve: `workload-2`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

### RDS PostgreSQL · Database writes

Estimated monthly service charges by Database writes. Adapter hosting and Tensor9 fees are excluded. Source: us-east-1; target: fr-par; currency: USD.

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/0d18fabe8236b2b263216a99e8aa472d282b6fd43d4fd16988c49e146c7f2e34-light.svg)

![RDS PostgreSQL · Database writes: monthly service charges by Database writes. Exact values are in the following table.](/images/diagrams/service-cost/c55e2766c185fb00263f1d1b108729664d0648da3cc24e7923f9532a816b4b81-dark.svg)

Comparison pills describe all sampled scale points; card amounts show the named workload. Costs The Same means within 20% at every point. Costs Depend includes mixed results and unpriced samples.

Exact sampled model outputs in USD/month. Difference means target minus origin; a positive difference means the target costs more.
Estimated allocation from planning assumptions; not benchmark calibrated. Lines connect evaluated samples; intermediate capacity and prices are not inferred. Unpriced means no target quote or cost difference is available within the sizing policy.

| Database writes (per second) | RDS PostgreSQL | Target scenario | Scaleway Managed Database for PostgreSQL | Difference |
| -: | -: | - | -: | -: |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 0 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 1 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 10 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 100 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 100 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 1000 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 1000 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |
| 10000 | 0 | Scaleway Managed Database for PostgreSQL (one copy) | 101.102715 | 101.102715 |
| 10000 | 0 | Scaleway Managed Database for PostgreSQL (two copies) | 154.68715395 | 154.68715395 |

**Estimated allocation from planning assumptions; not benchmark calibrated.**

The sizing policy grows compute with translated work and peak headroom, memory with compute and working set, and disk with resident working state and reserve. Monthly usage is distinct from resident data; each workload's assumptions define the translation. HA changes pricing, not this policy's throughput estimate. Unpriced workloads exceed the policy's configuration range; no silent vertical cap or automatic sharding is assumed.

| Sample / scenario | Basis | Mean / peak work/s | Required vCPU | Allocated vCPU | Memory GiB | SSD GiB | Resident GiB | Policy | Constraints |
| - | - | -: | -: | -: | -: | -: | -: | - | - |
| 0 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 0 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 0 / 0 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 1 / 2 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 10 / 20 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 100 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 100 / 200 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 1000 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 1000 / 2000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for PostgreSQL (one copy) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |
| 10000 per second / Scaleway Managed Database for PostgreSQL (two copies) | PlanningAssumption | 10000 / 20000 | 0 | 2 | 8 | 100 | 0 | `service-adapter-planning-2026-10-22.1` | None reported |

* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.

**Curve assumptions**

| Target scenario | Assumption | Value | Unit |
| - | - | -: | - |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (one copy) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (one copy) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Additional full-price standby copy | 0 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (one copy) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (one copy) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (one copy) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (one copy) | Dependent object sink write price guess | 0.000005 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary compute | 0.089 | USD/vCPU-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary memory | 0 | USD/GiB-hours |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Primary provisioned storage | 0.115 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database reads | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Origin list price · Database writes | 0 | USD/operations |
| Scaleway Managed Database for PostgreSQL (two copies) | Target CPU / source CPU | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target RAM / source RAM | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target disk / source disk | 1 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · read | 0 | USD/read |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · write | 0 | USD/write |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · provider operation | 0 | USD/provider operation |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-month | 0.11126565 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · published GiB | 0 | USD/published GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · delivered GiB | 0 | USD/delivered GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · network GiB | 0 | USD/network GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · replica GiB-month | 0 | USD/replica GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · vCPU-hour | 0.0367275 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target list price · GiB-hour | 0.006225 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-hour | 0 | USD/resource-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · resource-month | 0 | USD/resource-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · input token | 0 | USD/input token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target price guess · output token | 0 | USD/output token |
| Scaleway Managed Database for PostgreSQL (two copies) | Target monthly base charge | 0 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Target service minimum | 89.97615 | USD/month |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary CPU | 2 | vCPU |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary RAM | 8 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Minimum primary disk | 100 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Additional full-price standby copy | 1 | 0 or 1 |
| Scaleway Managed Database for PostgreSQL (two copies) | Allocated share of provisioned cost | 1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Read/write work per vCPU-second | 1500 | operations/vCPU-second |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning peak / mean work | 2 | multiplier |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning primary CPU utilization | 0.7 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Resident state reserved in RAM | 0.1 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Planning disk utilization | 0.8 | fraction |
| Scaleway Managed Database for PostgreSQL (two copies) | Maintenance and log disk reserve | 20 | GiB |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database CPU unit-price guess | 0.0413 | USD/vCPU-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database RAM unit-price guess | 0.007 | USD/GiB-hour |
| Scaleway Managed Database for PostgreSQL (two copies) | Staging database disk unit-price guess | 0.17 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink storage price guess | 0.023 | USD/GiB-month |
| Scaleway Managed Database for PostgreSQL (two copies) | Dependent object sink write price guess | 0.000005 | USD/write |

* Source compute and stored data remain independent; a throughput view needs explicit workload inputs. Editable authored planning guesses, not verified provider tariffs, benchmark measurements or a commercial quote. Other native meters retain their explicitly displayed default quantities.
* Planning estimate · editable authored rates and sizing assumptions dated 2026-10-22; not verified provider prices. us-east-1 → fr-par, USD/month.
* Source prices are reference charges, not an invoice. Target lines cover selected service usage/configuration only; adapter hosting and Tensor9 fees are excluded.
* Explicit tariff samples are not performance measurements or capacity qualification. No interpolation or break-even guarantee is implied.
* Variant assumptions show the first axis sample; any axis-dependent configuration is described by this curve's workload note.
* Rates are cited provider list prices, editable for discounts, commitments or another region. Workload translation and capacity coefficients remain editable planning assumptions, not benchmark measurements or a commercial quote.
* Default-zero rates are explicit assumptions, not evidence that the deployed service or resource is free.
* Provisioned-unit minimums use an editable approximate allocation floor. Modeled target service cost is max(usage/allocation, minimum), with billable network, durable staging and dependent sinks added separately; actual RU/TU sizing, scaling increments and throughput remain unqualified.
* Durable staging is priced once: provisioned targets include adapter state in their selected allocation; native targets add a separate staging allocation when required. Adapter hosting is a separate allocation and must exclude resources already priced as staging, workers or target capacity.
* A declared adapter mapping and cost model do not establish implementation availability or service equivalence.
* This is a declared directed mapping; a planning estimate does not establish deployment or request-serving support.
* Native/deployment comparison scope: This mapping provisions a native database. SQL clients connect directly to it; this directed mapping does not expose an RDS management adapter endpoint. Database availability and query latency are covered by the database provider, not a Tensor9 request-processing SLA.
* Authored planning assumptions, not measured adapter capacity. Origin rates are cited first-tier list prices, before free allowances and discounts.
* A cost scenario does not establish backend availability, protocol fidelity or production readiness. Registry serving status remains independent.
* Provisioned compute and RAM are priced by resource-hour; operations are independently authored. Database size cannot be inferred from dollars alone.
* Aurora/AlloyDB, MySQL/PostgreSQL and proxied targets retain their distinct backend contracts. HA, replicas and proxy allocations must be selected explicitly.
* RDS memory is included in the instance-hour price; the memory meter is zero so RAM is not counted twice.
* Scaleway Managed Database for PostgreSQL at Production Optimized (PRO2) prices in EUR converted at the ECB reference rate. A high-availability or replica node bills Scaleway's 'Additional nodes' price, EUR 0.0583 per hour for PRO2-XXS, 53% of the main node. Scaleway lists no egress charge for Managed Databases, so selected network is not charged. Cost Optimized ranges (DB-DEV-S from EUR 0.0156 per hour), Block Storage 15K and backups (EUR 0.03 per GB-month) are not selected.
* Row-store throughput (capacityOpsPerVcpuSecond, default 1500) is measured-by-tensor9 as a lower bound: the DynamoDB adapter on Cloud SQL db-custom-2-7680 (2 vCPU, PostgreSQL 16, us-west1) served 2,974 ops/s of YCSB-A at 16 threads without recorded saturation (src/main/kotlin/t9/prime/equiv/generated/eval/CloudsqlPostgresqlYcsbAT16GoYcsbV1\_0\_0.kt). Published CPU-bound pgbench reaches 3,101 read-write transactions and 20,289 point reads per vCPU-second (AWS z1d.2xlarge, PostgreSQL 14.5), so actual capacity is likely higher. Reads and writes share one coefficient: at the same adapter version, read-only YCSB-C was not faster than 50%-update YCSB-A.
* Log-broker throughput (capacityStreamMiBPerVcpuSecond 20 at 0.5 utilization, 4 GiB RAM per vCPU) is provider-documented by Google Managed Service for Apache Kafka sizing ([https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size](https://docs.cloud.google.com/managed-service-for-apache-kafka/docs/plan-cluster-size)): write-equivalent bytes are replicated writes plus a quarter of reads. Producer batches below 10 KB lower per-CPU throughput; that penalty is not modeled.
* Disk utilization 0.8 is provider-documented headroom below Cloud SQL automatic storage increase, which triggers near 91-95% full ([https://docs.cloud.google.com/sql/docs/postgres/instance-settings](https://docs.cloud.google.com/sql/docs/postgres/instance-settings)). The 20 GiB reserve exceeds PostgreSQL's 1 GB default max\_wal\_size; the remainder of that reserve is unmeasured.
* CPU tiers, 4 GiB per vCPU, at most 6.5 GiB per vCPU, 96 vCPU and 65,536 GiB disk follow Cloud SQL Enterprise custom machine limits (provider-documented, same page); 624 GiB is 96 × 6.5. Targets billed per node, unit or throughput unit are sized by their backend instead.
* No public evidence: peak/mean work (2) and the RAM working-set fraction (0.1). Proposed measurements: the p99 one-minute rate over the monthly mean from customer CloudWatch metrics; the buffer hit ratio while stepping RAM under a fixed YCSB-B load.
* Stored GiB already includes the origin family's physical footprint assumptions; no second storage amplification is applied.
* The workload family already selected resource-hours; the throughput coefficient is not applied again.
* Copies and allocation fractions change cost only; they do not increase modeled primary capacity.
* AWS Price List API, AmazonRDS offer: [https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json)
* Scaleway Managed Databases pricing: [https://www.scaleway.com/en/pricing/managed-databases/](https://www.scaleway.com/en/pricing/managed-databases/)

Cost scope: Modeled target native service and explicitly modeled state allocation; hosting and Tensor9 fees only when supplied.

Excluded costs and workload features:

* Additional network transfer, backups, KMS, one-time migration and operational labor unless an explicit meter includes them.
* Destination application execution and separately billed downstream services; supply those allocations separately.
* Aurora per-IO, serverless-capacity, backup and proxy tariffs; reserved/committed-use purchase options.
* Taxes, discounts, credits and free-tier allowances unless explicitly reflected in edited rates.
* Unmodeled dependent services, task execution and operations outside the selected workload family.

List prices read: 2026-10-22. Model: `plan-c828ea2296c34e1fe5b1f54217e1ed4e0bda56ff` / `2026-10-22.planning.2`. Curve: `workload-3`.

Sources: [AWS Price List API, AmazonRDS offer](https://pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonRDS/current/us-east-1/index.json), [Scaleway Managed Databases pricing](https://www.scaleway.com/en/pricing/managed-databases/).

See the [service catalog](/byoc/service-adapters/catalog.md) for supported directions, operation coverage and adapter limitations.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.