Removing abandoned restore points

[squebel]

I searched for a while but was unable to find exactly what I'm looking for so I here I am...

I'm seeing a lot of, what I'll call abandoned restore points out in our SOBR extents. One of the jobs I'm looking at is a Forever Forward Incremental that is set to keep 7 restore points. The chains look to be intact for the past 7 days, but I'm seeing several older vib files for each vm that are older than 10 days. I used a powershell script to pull a report and I got back an alarming number of files and they seem to have occurred on similar days. As far as I can tell, everything in this job is set up correctly and has been running fine.

My question is: Is there any kind of clean/automated way to have Veeam go through and validate old restore point files that are no longer used or do I have to do through some kind of manual cleanup process?

[Egor Yakovlev]

Hi Squebel,
each time job finishes, Veeam will check and apply specified Retention Policy. If Backup chain is set to Forever Forward Incremental(aka have NO Active\Synthetic full enabled), then on a disk you should see 1xFull(vbk)+6 increments(vib) at all times. Do you also see those "abandoned" restore points in VBR UI under Home-Backups-Disk?
Thanks!

[squebel]

Nope, they do not show up in the UI hence why I called them "abandoned." For example on one vm, there is the vbk followed by 6 vibs all in succession but then there are 3 other vibs from random dates that are older than oldest vib in the chain. What would cause this? I'm seeing this in all of my extents.

[Egor Yakovlev]

Incremental backup cannot exist without Full, and no matter what, restore points shall not disappear from UI without user intervention, so it is really worth investigating with support.
Please open a ticket and share its ID in this thread.

[squebel]

I can do that. However, I'm asking if there is a way to clean this up automatically or is it a manual method? By my calculations, I'm chewing up at least 18TB in abandoned vib files across my extents just in job folders that are 7 days FFI. I can't be the first person that's run into this issue or a similar issue.

[squebel]

Case 03935857 opened earlier today with Support and waiting to hear what they say...

[PeterC]

We see this also regularly on the SOBR extents. I always thought this would happen when a job failed for some reason.
It are indeed orphaned VIB files and they don't show up in the UI. We are using a script to clean up files older than a specific date.
I am curious to find out what will be the reason for this.

[stephen0830]

Can share the script to clean up the older Vin files?

[rossthorne]

I am running a SOBR with 10+ extents and have seen this happen quite often. I noticed this most when we would have a job interrupted when running a synthetic operation or when the job has an unexpected failure. I have talked with our Veeam reps and support in the past and the answer to this point has been to proceed as you are and remove anything that you know is outside your current retention and is older. I have had to and continue to do this manually because of how my jobs are staggered, I just don't trust a script(even if I write it) to not delete something that I might actually need. So I go to the folder, sort by create date, then compare that to the Oldest Full backup in the backup chain for each job.

I know your pain as I ran into this a while back as a first time SOBR user and we were using 120TB or almost 10% on orphaned files.

[Justus]

A caution for anyone manually removing files. According to Veeam tech support, the file system dates are not a reliable way to check for files last used or created. You need to use the date embedded in the file name.

[socra]

I can do that. However, I'm asking if there is a way to clean this up automatically or is it a manual method? By my calculations, I'm chewing up at least 18TB in abandoned vib files across my extents just in job folders that are 7 days FFI. I can't be the first person that's run into this issue or a similar issue.

@squebel
Curious to learn what support comes up with, please keep us posted on this.
Thanks.

[squebel]

At this point, Support has been of no assistance. They keep trying to find the problem but my ask of them is how to clean it up first. To this point, they have provided no suggestion on how to do that officially.

Thank you to the others following this thread that have said you're having the same problem. It seems like Veeam needs to find the cause of the problem and provide an official clean-up solution.

Thanks @Justus for the suggestion about the date in the filename. However, I think the last modified times for the vib files should be able to used. For example, the jobs that are showing the largest amount of abandoned files are only 7 day jobs. Anything with last modified times older than say 10 days in the those folders should easily be able to be trashed without causing an issue. I have noticed that you can't go by created time though because it will show the vbk's since they show the original date of the first Full as the creation time.

[Sylvala]

Thank @squebel. Veeam miss the feature to do a job just for managment and clean-up and optimize without doing actually a backup job or else.

[squebel]

As suspected and confirmed with Support, there is no method for cleaning this up... you're on your own.

[GianlucaCroci]

perhaps this can help...

set _Dest=drive:\folder\subfolder
set _olderthan=30
FORFILES /P %_Dest% /S /D -%_olderthan% /C "cmd /c del /s /q @file"