SOBR with linux hardened repositories

[StanoSedliak]

Hi,

I would have a question about how SOBR is working in this scenario, if someone please would have some hints how it working please.

We have hardened linux repositories added as performance tiers to a SOBR. The backup jobs have a synthetic full (before I had jobs on a standalone hardened repo where reflink feature from the linux repo with XFS filesystem saved us a lot of space).
Now I have a SOBR and I hope again reflink will save as a lot of space as we have synthetic full turned on, but my question is if for example we have repo1 and repo2 in the SOBR and backup job for customerA is running on repo1, synthetic full is scheduled and will run, reflink will save us a lot of space but if there will be no space left on repo1 the next full backup job for customerA will be placed on other repo for example repo2 so there will be no saving on space as reflink can not point to existing "block" from repo1?

How to operate backups located on SOBR with hardened linux repositories to use the reflink technology and save space?
Thank you!

[haslund]

You can't gain reflink (fast cloning) savings across different disks or repositories. Veeam will do a great job of splitting out the backup jobs and chains on different extents within the SOBR, but yes if one repo is running out of space you will either need to extend the capacity, let it create a new chain on a different extent or decrease retention.

[StanoSedliak]

Hi Rasmus,

thank you for you answer.

So basically in this case if we will use SOBR we will lose the advantage of the reflink/space saving. The only way if we need to keep space saving because full jobs are really big by us, we need to use hardened repo as standalone veeam repositories and if there will be some problems with space we need to use move backup to another box etc... but we will save a lot of TB on the backup storage.

[StanoSedliak]

Some additional question.
If I will have several standalone hardened linux backup repositories and for example repo01 will run almost out of storage so I will move in the reality just copy existing backup for customerA inside Veeam to another repo02, it will copy the backups as on repo01 is still the immutab. flag, after this is done, I will go to repo01 change the VBK and VIB files manually from immutable to non immutable, but how is the correct procedure in Veeam to be able afterwards make space on the repo01? will the backups be shown in orphaned and can I than just delete it if I will put manually the immutable flag away?

Thank you!

[bct44]

From my POV you can use sobr and use the functionnality evacuate and put your extent in maintenance. You won't lose space saving from block cloning, specially if you are running your backup in true-per vm backup files.
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

For your additional question, you can delete files manually as you described. Don't forget to rescan your sobr/repo after the deletion action, you will probably need to forget all unavailable backup in the db.
https://helpcenter.veeam.com/docs/backu ... ml?ver=120