Backup failures with "Unknown backup format VBK" error

[vmmatty]

I have my backup jobs split into two jobs that run at the same time. Since there is a very large number of VMs this lets me finish the backups on time. One of the jobs has been working great since I set it up. The other one backed up for two nights and now errors every night with the following error for every VM it tries to backup:

Checking backup consistence
Reading backup directory failed: Unknown backup format VBK ERROR:Storage [Block offset refers to area outside the storage. Storage size: [325091723855]. Block offset: [3891110085
161648128].] is corrupted.

Does this mean my backup file on disk has become corrupted and I need to delete it? That will mean I've lost all the rollbacks in that backup job. Any other suggestions?

[Gostev]

Matt, there could be different causes for this error (not just actual backup file corruption), best would be to investigate the specific issue with full logs, so please open a case with our support. In all cases, we have some support tools which allow for recovering most of the data from backup file as well as rollbacks even in case of actual hard disk corruption.

[vmmatty]

There is no way to do any more troubleshooting without opening up a case? It wouldn't be the end of the world if I had to delete and recreate my backup files (since I just started using this for backups a week or so ago), but I wanted to know what is going on for future reference. Anything you can suggest without opening up a support case?

[Gostev]

Frankly this specific issue is too technical for me, I am not doing the actual support so I cannot suggest a lot of things about complex issues like this one. I am guessing that even support will be unable to troubleshoot this without seeing the logs.

What I would personally do is preserve all backup files in case need arises to restore from them, and just re-create the backup job. If you do not archive your backups to tape, I would also suggest implementing rotating media scenario, we are doing this internally and I know vast majority of our customers without tape devices are doing exactly the same. Hard disks are definitely not the most reliable things, so it is better to have insurance

[vmmatty]

I deleted the existing backup files and will let the jobs run again. If it gets corrupted again, I will contact support.

Thanks.