KnearMe Edge Worker Operations Runbook

Incident triage sequence

1. Check Worker health endpoint (/_health).
2. Identify affected route set and current ownership stage.
3. Inspect queue consumer and DLQ activity.
4. Verify upstream origin availability.

Fast rollback

Route-based rollback is the fastest mitigation:

1. Disable ownership for affected path(s).
2. Confirm traffic returns to origin.
3. Keep or pause queue consumers based on incident scope.

Verify protections

• Rate limit behavior (429 with Retry-After).
• Turnstile enforcement (403 on invalid token if enabled).
• Redirect sync path admin auth.

Post-incident tasks

• Capture route stage at failure time.
• Record queue depth, DLQ count, and request class impacted.
• Update cutover runbook notes for recurring failure pattern.