Backup copy job creates an unwanted full every day

[DanielJ]

We have a StoreOnce appliance since years. Trying to use it with VBR has been a neverending parade of failures. The current problem: a copy job with many small vms and 3 "big" ones (over 1 TB) failed to build synthetic fulls for the "big" vms since some time last fall. Support was on a level of "have you tried restarting the StoreOnce". No solution was found. Some weeks ago I switched the job to "read the entire restore point" to avoid synthetic operations. Now it doesn't fail any more, but it creates an active full every day for the 3 big vms. It looks just like it would if there was a daily GFS setting and we had it activated. The retention settings we have are 30 daily, 5 weekly, 12 monthly and 3 yearly. The customer's requirement of long-term backups makes me unwilling to just create a new job to see if that would help. The backup data for this copy job is now over 320 TB. Even though the job "succeeds" it has stopped deleting expired backups for the 3 big vms.

Support is struggling. So far I have got nothing besides requests for yet more logs and a copy of the database. Ticket: 07600224

From here it looks like the StoreOnce integration is a non-working mess. I would have used it as a simple NAS if we hadn't been required to have immutability on all backups.

[david.domask]

Hi Daniel,

Sorry to hear about the challenges and thank you for sharing the case number.

Based on the most recent notes, it looks like there is an unusual situation in the Veeam configuration database; while I understand no actionable progress has been made, the most recent update from Support looks to have identified that there is a check which is failing to find previous full backups, so the job is forcing Full backups.

This is of course not expected behavior, but from my perspective looks like Support has a good lead so far and they're working to understand how this situation occurred.

I'm not confident at first blush this is about the Storeonce itself, but I understand the sentiments. Let's give Support time to continue the investigation, but looks like they are following the steps leading up to this behavior fairly well in the case. I will ask them to provide an update with the current state of research.