Backup & Continuity
Backup Monitoring & Testing
Backup software reports success far more often than backup actually succeeds. The difference is only visible if somebody is looking.
We monitor every backup job and chase the failures, including the quiet ones: the server that dropped off the schedule, the database that was open at the time of the copy, the job reporting success against a target that has been empty for weeks.
On a schedule we restore from backup deliberately, into an isolated place, and confirm the data comes back usable rather than merely present. That is the only test that means anything.
You get the results either way. A month where something failed is considerably more useful to know about than a month where the report simply said it had not.
What this covers
- Daily monitoring of every backup job, with failures chased rather than logged.
- Scheduled test restores, performed rather than assumed.
- A check that what comes back is usable, not just present.
- Coverage gaps flagged when a new server or service appears.
- A plain summary of what is protected and how far back it goes.
Common questions
Our software says the backups are fine. Is that enough?
It is a good sign and not a proof. Backup software reports on whether the job ran, not on whether the data inside it would restore. Those two come apart more often than people expect, which is the entire reason the test restore exists.
How often should restores be tested?
Often enough that a problem surfaces inside a window you could live with. For most businesses that is a full test quarterly with lighter checks in between, and the right answer depends on how much data you could afford to lose.
Part of Backup & Business Continuity
Backup that somebody has restored from on purpose, so the first real test is not the day you need it.