| name | backup-recovery |
| description | 3-2-1 backup strategy with verified restore procedures. Use when setting up or reviewing backup infrastructure, before risky data migrations or destructive operations, or after adding new persistent storage such as a database or volume. |
| license | MIT |
Backup & Recovery
Ensure data safety through verified backup strategy.
When to Use
Protects and restores persistent data (databases, volumes) — 3-2-1, verified
restores. Its sibling rollback-plan covers the other kind of undo: reverting
a config/infra change to a prior known-good state. Data loss → backup-recovery;
a bad deploy or migration to walk back → rollback-plan (a risky migration often
wants both).
Procedure
- Identify targets: List all persistent data (DBs, volumes, configs).
- Apply 3-2-1 rule: 3 copies, 2 different media, 1 offsite.
- Automate: Every DB or persistent volume must have an automated backup job.
- Verify restore: A backup is not valid until a restore has been tested. Run periodic restore drills.
- Document: Record backup schedule, retention policy, and restore procedure in project docs.
Completion
- All persistent data has automated backup jobs
- Restore tested at least once
- Schedule, retention, and restore steps documented
Anti-Patterns
- Untested backups — a backup that can't restore is worthless
- Manual-only backups — will be forgotten and skipped
- No offsite copy — local disaster destroys all copies
