A backup you have never restored is a hypothesis
Job success reports measure whether the software finished. They do not measure whether your data comes back. Those are different questions.
By Vilindor Operations
Backup reporting is optimistic by design. A job completes, the dashboard turns green, and everyone moves on. What the green does not tell you is whether the files inside are consistent, whether the application will start from them, or whether the retention policy quietly aged out the copy you will eventually need.
The failure modes that pass as success
- Open files skipped silently, so the database inside the backup is not consistent
- Retention shortened during a storage squeeze and never reverted
- A new server added to the estate that was never added to a backup job
- Backups written to a share the domain admin account can delete, which ransomware will find
- Application-aware processing disabled during troubleshooting and left off
Test restores are the only real evidence
There is exactly one way to know that recovery works, and it is to recover something. Not annually as an audit exercise, but on a rolling schedule that samples across the estate — a file here, a full VM there, a mailbox somewhere else — and produces written evidence each time.
This is unglamorous work and it is the reason it gets skipped. It is also the difference between a recovery that takes four hours and one that turns into a business-ending week.
Immutability is now table stakes
Modern ransomware operators look for backups first, and they frequently find them because the backup repository is reachable with the same credentials as everything else. An immutable or genuinely air-gapped copy is what survives an attacker who already has domain admin. If you have not verified that at least one copy cannot be deleted by a compromised administrator account, that is the most valuable hour you will spend this month.