Veeam target | How to detect CRC errors in time?

Question

A custumer uses Veeam and a NAS as primary backup target. The backups for last days were performed without any errors. Now they cannot restore due to CRC errors. Obviously, the writing to the NAS worked, but reading fails.

I wonder now how to prevent this in the future.

Veeam offers maintanance (backup files health check) for backup targets.

From the Veeam documentation:

An automatic health check allows you to avoid a situation when a restore point gets corrupted, making all further increments corrupted, too. If Veeam Backup & Replication detects corrupted data blocks in the restore point during the health check, it will transfer these data blocks to the target backup repository during the next backup copy interval and store them in the newly copied restore point.

The configuration dialog box also states that Maintenance is not required when periodic full backups are enabled.:

Questions:

• Does this mean that the integrity check is performed only for the incremental backups (.vib files) and full backups are not checked?
• Would CRC errors be detected while using the maintenance, even if only full backups are performed?

Answer 1

Maintenance checks are performed whenever you've set them to run.

The point is that forever-incremental with synthetic fulls could accumulate errors that stay in the backup as long as the original data isn't rewritten, causing them to be reread in the next incremental run.

With active fulls, data is read from the original on a regular basis, so any introduced error will vanish on the next active full run.

In either case, you need to ensure that your repository is reliable. For instance, a NAS with broken memory is likely to trash the data written to it. Basically, any backup written to it will be unreliable, no matter if active full or incremental.

Using synthetic instead of active fulls increases the chance for data corruption. Using long chains of incremental backups also increases the chance for data corruption. Running weekly synthetic fulls creates additional reference points and decreases corruption likelihood.

Essentially, if you don't know whether the repository works reliably you need to seriously test it or replace it.

Additionally, you need to test your backups on a regular basis - at least annually. A backup you can't trust has no worth.