Prometheus

Workflow

1. Confirm SLOs and critical service signals.
2. Define metric model and label policy before instrumentation.
3. Configure scrape jobs and service discovery with security constraints.
4. Implement recording rules for expensive or standardized queries.
5. Build alert rules with clear severity and runbook links.
6. Tune retention and storage based on ingestion profile.
7. Validate alert and query behavior under failure scenarios.

Preflight (Ask / Check First)

• Prometheus version and deployment model.
• Ingestion volume and retention targets.
• Alertmanager routing and escalation policy.
• Known cardinality hotspots.
• Multi-tenant or federation requirements.