DevOps Professional · Module 3: Migration, restore and disaster recovery · Lab
Diagnose a Broken Backup, Verify the Good One Separately
50 min hands-on · Advanced
A local machine with two fixture backups (one staged broken, one good) and a separate restore target. Small data standing in for scale; labeled as such.
Local guide: run the steps below on your own machine in order, then check the validation list.
Objectives
- Diagnose the broken backup from drill evidence
- Restore the good backup to a separate target timed end to end
- Write the drill log with findings and the next due date
Step 1
Break and diagnose
Run the restore drill against the broken backup, capture the failure, and diagnose its cause class from evidence (credentials, paths, format or runbook rot in this staging). Quote the diagnosis.
Step 2
Restore the good one separately
Restore the good backup to the separate target (never over the live fixture), time it end to end, and verify the data against the stated RPO with quoted checks.
Step 3
Log the drill
Write the drill log: what passed, what failed with its owned fix and re-drill date, the next due date, and the honesty paragraph on what small-scale drills do not prove.
How to confirm it worked
- Broken backup diagnosed with quoted evidence
- Good restore timed end to end on a separate target
- RPO verification quoted from restored data
- Drill log with findings, fixes and next due date