From d4a3b5e432afd7ab7469c98ccd2df9d9f308144a Mon Sep 17 00:00:00 2001 From: Durable Workflow Date: Fri, 4 Sep 2026 02:40:28 +0000 Subject: [PATCH 1/3] Publish Cloud Dev capacity benchmark --- docs/polyglot/cloud-control-plane.md | 43 ++++++++++++++++++++++++++++ 1 file changed, 43 insertions(+) diff --git a/docs/polyglot/cloud-control-plane.md b/docs/polyglot/cloud-control-plane.md index 3bb19c9ca1..753e84c5c3 100644 --- a/docs/polyglot/cloud-control-plane.md +++ b/docs/polyglot/cloud-control-plane.md @@ -235,6 +235,49 @@ and failover independently. A self-hosted Server cannot be registered with, attached to, or used as the backing runtime for a Cloud namespace. Embedded Laravel, self-hosted Server, and Cloud are separate deployment choices. +## Cloud Dev Capacity + +Cloud Dev is an isolated, single-host managed runtime for development and +evaluation. Each provisioned space receives the same runtime shape used for +the measurement below: + +| Resource | Cloud Dev | +| --- | --- | +| Runtime compute | 1 shared vCPU, 1 GB RAM | +| Durable storage included | 5 GB | +| Runtime services | Server, queue worker, scheduler, MySQL, and Redis | +| Network path | Direct, space-specific HTTPS ingress | +| Customer workers | Run in the customer's environment | +| Availability | No SLA; maintenance interruptions are allowed | +| Runtime price | $0.03/hour, measured by minute, capped at $20/month | +| Storage above 5 GB | $2/GB-month | + +Cloud Dev was measured with [DW Standard Workflow +v1](https://github.com/durable-workflow/server/tree/main/benchmarks/capacity), +a fixed comparison workload consisting of one workflow start, one external +activity, and one workflow completion with defined 1 KiB Avro inputs and +results. The test used the provisioned 1-vCPU/1-GB runtime topology, current +stable Server and PHP SDK artifacts, 32 worker processes, 16 client slots, a +60-second warmup, and a five-minute measurement window. + +| Measured result | Value | +| --- | ---: | +| Offered and completed rate | 0.10 standard workflows/second | +| Completed workflows | 30 of 30 | +| Errors / throttled starts | 0 / 0 | +| Scheduling latency, p50 | 30.5 ms | +| Scheduling latency, p95 | 307.1 ms | +| Scheduling latency, p99 | 827.0 ms | +| Final workflow backlog | 0 | +| 30-day equivalent at the measured rate | 259,200 standard workflows | +| 30-day workflow-start + activity equivalent | 518,400 operations | + +This is a measured development baseline, not a universal conversion for every +workflow and not an SLA. Larger payloads, additional activities, timers, +signals, queries, replay-heavy histories, and customer worker latency change +capacity. Cloud billing remains based on provisioned runtime time and durable +storage, not workflow operations. + ## Billing Usage API Cloud Dev uses one isolated, single-host runtime cell for each provisioned Dev From cbf898a81c1196f6ef3daccab7f7c31be4107472 Mon Sep 17 00:00:00 2001 From: Durable Workflow Date: Fri, 4 Sep 2026 02:41:24 +0000 Subject: [PATCH 2/3] Correct Cloud Dev benchmark headline --- docs/polyglot/cloud-control-plane.md | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/docs/polyglot/cloud-control-plane.md b/docs/polyglot/cloud-control-plane.md index 753e84c5c3..7f444ef432 100644 --- a/docs/polyglot/cloud-control-plane.md +++ b/docs/polyglot/cloud-control-plane.md @@ -257,20 +257,20 @@ v1](https://github.com/durable-workflow/server/tree/main/benchmarks/capacity), a fixed comparison workload consisting of one workflow start, one external activity, and one workflow completion with defined 1 KiB Avro inputs and results. The test used the provisioned 1-vCPU/1-GB runtime topology, current -stable Server and PHP SDK artifacts, 32 worker processes, 16 client slots, a -60-second warmup, and a five-minute measurement window. +stable Server and PHP SDK artifacts, one PHP worker process, two client slots, +a 30-second warmup, and a five-minute measurement window. | Measured result | Value | | --- | ---: | -| Offered and completed rate | 0.10 standard workflows/second | -| Completed workflows | 30 of 30 | +| Offered and completed rate | 0.25 standard workflows/second | +| Completed workflows | 75 of 75 | | Errors / throttled starts | 0 / 0 | -| Scheduling latency, p50 | 30.5 ms | -| Scheduling latency, p95 | 307.1 ms | -| Scheduling latency, p99 | 827.0 ms | +| Scheduling latency, p50 | 28.0 ms | +| Scheduling latency, p95 | 95.1 ms | +| Scheduling latency, p99 | 134.5 ms | | Final workflow backlog | 0 | -| 30-day equivalent at the measured rate | 259,200 standard workflows | -| 30-day workflow-start + activity equivalent | 518,400 operations | +| 30-day equivalent at the measured rate | 648,000 standard workflows | +| 30-day workflow-start + activity equivalent | 1,296,000 operations | This is a measured development baseline, not a universal conversion for every workflow and not an SLA. Larger payloads, additional activities, timers, From d4d8d01064fa89f2c5db60fe203b748ae8a88666 Mon Sep 17 00:00:00 2001 From: Durable Workflow Date: Fri, 4 Sep 2026 02:42:52 +0000 Subject: [PATCH 3/3] Use workflow actions for monthly comparison --- docs/polyglot/cloud-control-plane.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/docs/polyglot/cloud-control-plane.md b/docs/polyglot/cloud-control-plane.md index 7f444ef432..e97fb642b4 100644 --- a/docs/polyglot/cloud-control-plane.md +++ b/docs/polyglot/cloud-control-plane.md @@ -269,8 +269,7 @@ a 30-second warmup, and a five-minute measurement window. | Scheduling latency, p95 | 95.1 ms | | Scheduling latency, p99 | 134.5 ms | | Final workflow backlog | 0 | -| 30-day equivalent at the measured rate | 648,000 standard workflows | -| 30-day workflow-start + activity equivalent | 1,296,000 operations | +| 30-day workflow actions | 1,296,000 | This is a measured development baseline, not a universal conversion for every workflow and not an SLA. Larger payloads, additional activities, timers,