Incident Triage And Oncall Runbooks

Use When

• Service health degradation is detected
• User-impacting errors or latency spikes are active

Workflow

1. Classify incident severity and blast radius.
2. Gather evidence from dashboards, logs, traces, and recent releases.
3. Decide mitigation path (rollback, feature flag disable, traffic shift, hotfix).
4. Communicate timeline, owner, and next update checkpoint.
5. Capture post-incident actions and preventive fixes.

Required Checks