Waiting for live workflow status signals.
Observed workflow conditions, not confirmed customer outages. Owners are catalog assignments.
Open the workflow source and recent history. Confirm customer impact and the latest observation time before escalating. A recorded owner is a catalog assignment, not a confirmed on-call responder.
For workflow failures, contact the recorded owner with the workflow ID, failed run, timestamp, source link, and customer impact. If no accountable owner is recorded, obtain an engineering incident owner before handing off.
For monitoring problems or stale observations, inspect the status publisher and its CloudWatch logs first. Missing telemetry does not prove that the underlying service is down.
CloudWatch alarms and dashboard conditions do not confirm that anyone was paged. Verify the alarm’s notification actions and the PagerDuty incident before relying on automated escalation. Use your established incident escalation channel for urgent impact.
| Workflow | Path | Role | Status | Signal | Freshness | Infra |
|---|