Runbooks

Runbooks should be direct, current, and written for people under pressure.

Template

Use this structure for each runbook:

• Symptom
• Impact
• First checks
• Recovery steps
• Escalation path
• Related dashboards, alerts, and logs

Initial Runbooks

• Local PostgreSQL
• Local pgBackRest
• Local Airflow
• Local dbt
• Local Great Expectations
• Local Superset
• Local OpenMetadata
• Local CKAN
• Local OPA
• Local OpenBao
• Local MinIO
• Extract and Load Connectors
• M365 Graph Setup
• Superset M365 User Memberships
• Superset Naming
• Failed data pipeline
• Delayed source extract
• Schema drift detected
• Access request blocked
• Production deployment rollback