Skip to content

perf(worker): cap Fargate concurrency at 10 - #287

Merged
suguanYang merged 1 commit into
mainfrom
perf/wangbinqi/worker-concurrency-10
Aug 14, 2026
Merged

suguanYang merged 1 commit into
mainfrom
perf/wangbinqi/worker-concurrency-10

Conversation

@suguanYang

Copy link
Copy Markdown
Contributor

Summary

  • set WORKER_CONCURRENCY=10 explicitly in the staging Fargate task definition
  • extend the ECS rendering contract so the concurrency target cannot drift with the application default

Rationale

The previous staging task omitted this environment value and inherited gevent concurrency 50. Under the prior pressure run, the single 0.5-vCPU worker saturated and lost health. This change establishes the owner-approved per-task target before the controlled capacity re-baseline.

Validation

  • uv run pytest deploy/ecs/test_render_task_definitions.py — 6 passed
  • uv run --group lint ruff check deploy/ecs/render_task_definitions.py deploy/ecs/test_render_task_definitions.py — passed
  • full repository suite — 472 passed; 4 unrelated existing failures in summary-builder/page-memory tests

Supports Ontos-AI/knowhere-api-infra#22.

@suguanYang
suguanYang merged commit 2e4eb58 into main Aug 14, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant