Error monitoring / observability - self-hosted Sentry-alt + AWS logs + Slack alerts + retries¶
| Stage | Estimated |
|---|---|
| Status | Deferred |
| Design status | In progress |
| Estimate | 1w (M) |
| Confidence | Medium |
| Linear | PIN-332 ↗ |
| Linear status | In Progress |
| Design | Design ↗ |
| Linear epic | Platform Foundation |
| Module | Settings & Admin ↗ |
Priority: Medium
Scope¶
Production observability + safe-deploy pipeline: blue-green CI/CD deploys gated on health checks, Sentry wired into both the API and frontend, OpenTelemetry traces flowing to the production backend, and a documented rollback runbook. Error alerts route to Slack with an on-call rotation. (Source: PIN-332.)
Acceptance criteria¶
- Blue-green deploy pipeline: new revision must pass health checks before traffic switches to it
- Sentry error tracking live in production for both the Go API and the web frontend
- OpenTelemetry spans flowing to the production observability backend
- Slack alert channel fires on production errors
- Documented rollback runbook: revert traffic, roll back DB migrations, incident-communication template
- On-call rotation documented with Sev1 escalation contacts
Conor's comments¶
Who is this assinged to?
Notes¶
Slack pod-crash alerts live; connect self-hosted monitor + AWS logs; add client-side retries
Open question for Conor¶
-