Metrics Troubleshooting
Diagnose empty metrics, stale health checks, missing services, and advisor result issues.
Diagnose empty metrics, stale health checks, missing services, and advisor result issues.
Coverage
| Item | Status |
|---|---|
| Check table existence | Use deployed route documentation |
| Check cron/worker state | Use deployed route documentation |
| Check token role | Use deployed route documentation |
| Check polling frequency | Use deployed route documentation |
Workflow
- Confirm each route against the deployed route documentation.
- Test with a staging token first.
- Use read endpoints to inspect current state before any write operation.
- Record the external workflow ID in your own system.
- Review jobs, request logs, and audit logs after the workflow runs.
Notes
- Do not create production automation around undocumented routes.
- Use idempotency keys for write endpoints.
- Keep rollback steps documented with the integration runbook.
Diagram
flowchart TD
A[External System] --> B[Read Current State]
B --> C{Expected State?}
C -- No --> D[Stop and Alert]
C -- Yes --> E[Run Approved Workflow]
E --> F[Read Jobs and Logs]
F --> G[Operator Review]