1

Hi

I have a job with reverse incremental the full is incomplete so no restore, I get All instances of the storage metadata are corrupted. I see thath there are 20 roll backs all ok but the full is not complete. Can the roll backs be used to repair the full?

Many thanks for any help

3 Spice ups
2

It’s probably best to open a case with Veeam, however if your full is corrupt or incomplete, restores may not be possible as this is the base required for the incremental.

You may have to create a new full and start again.

Negative.

1 Spice up
3

What software (version & edition) are you using ?

“Veeam” have 28 different products so it is hard to guess which you might be using ?

But in short, the full must be first completed. If full is not complete, the jobs would have a fail status…then nobody rectified ?

1 Spice up
4

Hi @BBBricks, and welcome! You’re already getting some good advice, but I wanted to add a few things to think about.

The problem with Reverse Incrementals is that each job injects recent changes into the most recent full. The benefit of that (historically) was that your last Full backup was always up to date and ready to restore if need be. There was not a need to first restore the last Full backup and then each of the Incrementals to get a server up to date. Reverse Incrementals are not as useful anymore because most backup software (including Veeam B&R) allows you to simply select the last restore point (Full or Incremental) and the software will figure out the rest to complete the Restore job.

The other problem with Reverse Incrementals is that you are constantly opening and writing to the same backup file. This increases the chance of that file becoming corrupted. I have seen that happen multiple times.

Every environment is different with different requirements for Backups and Restores, so coming up with the best scheme for your systems depends on that. For most of my clients (small businesses) a nightly Incremental and a weekly Synthetic Full is plenty. If you have a system that is highly transactional, you might need to run backups more often. You can also set up multiple jobs for the same server/system. Maybe have a regular nightly Incremental so you have a backup chain that corresponds to your document retention policy and run a Reverse Incremental on a more frequent schedule so you always have an “Oh Crap!” copy up to date.

Just some stuff to ponder. :slight_smile:

3 Spice ups