Almost every business we talk to says they have backups. Far fewer can say when they last restored something from them. That gap matters, because a backup that has silently stopped running, skipped a folder or cannot be decrypted looks exactly like a working one until the day you need it.
A restore test does not have to be complicated. An hour or two every quarter is enough for most small offices.
Why backups fail quietly
- A job stopped months ago after a password change, a full disk or an expired subscription, and nobody was watching the alerts.
- New data was never added. A new shared folder, a new server or a new laptop was set up after the backup was configured.
- Cloud apps were assumed to be covered. Microsoft 365 and Google Workspace keep deleted items for a limited time, but that is not the same as an independent backup you control.
- The encryption key or password is lost. The backup exists but nobody can open it.
- Ransomware reached the backups too. Backups stored on a drive or share that is always connected can be encrypted along with everything else.
A quarterly restore drill
- List what matters. Shared drives, accounting data, line-of-business databases, email, OneDrive or SharePoint, and any important files that only live on individual laptops.
- Check the reports. Confirm every item on that list had a successful backup in the last day or so, and that alerts go to someone who reads them.
- Restore a few individual files from different locations and different dates, to a separate folder, and open them to make sure they are usable.
- Restore something big. Once or twice a year, restore a whole server, virtual machine or database to a test location. Time how long it takes.
- Restore from the off-site copy, not just the local one. If the office were unavailable, this is the copy you would be relying on.
- Write down the result: what was tested, how long it took and anything that did not work. Fix the problems before the next drill.
The 3-2-1 rule, updated
A good starting point is three copies of your data, on two different types of storage, with one copy off-site. Today we also recommend that at least one copy is immutable or offline, meaning it cannot be changed or deleted for a set period, even by an administrator account. That is what lets you recover after ransomware rather than paying for it.
How long can you afford to be down?
Restore testing also answers a business question: if the server failed this morning, how long until people could work again, and how much recent work could you afford to lose? If the honest answer from your drill is "two days" and the business can only cope with two hours, you need a different approach, such as local image backups for fast recovery or a standby system in the cloud.
Getting help
Our managed backup service monitors every job, alerts us when something fails and includes scheduled restore tests. We also design backup and disaster recovery plans and business continuity for organizations across Los Angeles County, the Inland Empire and Ventura County. If a restore has already failed, our data recovery team can help.