Before starting
- Which system, failure and customer task does this cover?
- Who can authorize the action and who must be informed?
- Which access, backups and dependencies are required?
- What conditions mean stop and seek help?
Resources / Recovery planning
Could another responsible person follow the recovery steps?
Use these questions to prepare a runbook and an agreed exercise. This checklist does not authorize a production test.
During recovery
- Are steps ordered with expected observations?
- Which actions change or remove data?
- How are decisions and times recorded safely?
- What is the fallback if a step fails?
After restoration
- Can the important task be completed again?
- Are records consistent and deferred jobs accounted for?
- Have monitoring and alerts returned to the expected state?
- Who owns follow-up work and the next review?
Keep the exercise bounded.
Choose an isolated scenario or approved operating window with explicit limits. Preserve recovery material until the agreed checks pass. A configured backup job is not proof that the necessary data can be restored.