Return dependency readiness from the public API health endpoint
Deploy / deploy-dev (push) Skipped
Deploy / deploy-prod (push) Successful in 2m43s

This commit is contained in:
Corey Blais committed 2026-10-22 09:31:25 -04:00
1 parent 474b54e6ec
commit 138ea78ec7
4 files changed
+75 -19

No files matched your search

+3 -2
View File
@@ -55,8 +55,9 @@ Monitor these production checks:
- Verifies Nginx is serving the frontend container.
- Backend liveness: `GET https://your-host/api/health/live`
- Verifies the API process is running.
- Backend readiness: `GET https://your-host/api/health/ready`
- Verifies the API can reach Postgres and Redis. Returns `503` if either dependency is unavailable.
- Application health: `GET https://your-host/api/health` (alias: `/api/health/ready`)
- Use this endpoint for application availability monitoring. Verifies Postgres and Redis, returning HTTP `503` if a check fails or exceeds four seconds, and HTTP `200` only when all checks pass. Responses are not cacheable.
- Connection failures and timeouts also mean unavailable: during a startup database failure, the API may not be listening at all.
- Backend metrics: `GET https://your-host/api/metrics`
- Admin-authenticated process, request, and queue metrics.
- Postgres and Redis: