What you will achieve
Prove your backups actually work by performing a controlled restore drill — recovering one file, one folder, or booting a recovery environment without waiting for a disaster.
1) Plan the drill scope
- Pick one backup type you rely on: OneDrive, File History, system image, or third-party tool.
- Choose a non-critical test file — or a copy of an important document with a known checksum.
- Document start time, steps, and result in a simple log (date, backup source, success/fail).
2) File-level restore test
- Delete or rename a test file in Documents.
- Restore from OneDrive version history, File History (Settings → Update & Security → Backup → More options → Restore files), or your backup app.
- Open the file and confirm content matches the original.
3) System image drill (quarterly)
- Boot the recovery USB without touching the production disk if possible — use a spare VM or old laptop.
- Navigate to System Image Recovery and confirm Windows sees the latest image on external storage.
- Cancel before restoring if this is a detection-only drill — the goal is verifying the image is readable.
4) Schedule the next drill
- Calendar reminder every 3–6 months.
- After major PC changes (new disk, clean install, encryption), run an extra drill within a week.
5) Document RTO/RPO targets
- Recovery Time Objective: how long you can be down. Recovery Point Objective: how much data loss is acceptable.
- Your drill should prove you meet both — if restore takes 8 hours but RTO is 4, fix the backup strategy.
6) Offline media test
- Disconnect backup drive from network (simulate ransomware) and confirm you can still read backups from a clean PC.
7) Tabletop exercise checklist
- Who restores? Which machine? Where are keys/passwords? Time the full process — update runbook with actual minutes taken.
8) Partial restore scenarios
- Practice restoring single mailbox, single DB table, and full VM — different RTO for each tier.
- Log defects when restore fails — fix backup config before real incident.
Verification checklist
Record actual restore minutes in runbook. If twice target RTO, upgrade backup method or document accepted risk for management sign-off.
- Reboot once after changes that affect services, drivers, or firmware.
- Confirm the original problem is resolved under normal daily use, not only immediately after the fix.
- Note date, Windows version (Settings → System → About), and what changed in your personal runbook for next time.