Backup VM using volume groups: no expiration date for Recovery Points?

[PAP400]

Hello evb,

I have a couple of VM that use "Volume Groups" as storage.

A daily classic Veeam Backup Job create a VM snapshot, but also its Volume Group Recovery Point.
While the VM snapshot is automatically deleted and only the last one remains, the "Volume Group Recovery Point" is created without an expiration date and therefore it remains on storage pool for 60 days (fortunately AOS marks them to last for 60 days since 5.18)

I know that
"...Veeam Backup & Replication takes VG snapshots only if the backup scope includes individual virtual machines with volume groups attached...."
"...VG snapshots allow Veeam Backup & Replication to use the CBT mechanism while creating backups and to restore VMs with volume groups..."
https://helpcenter.veeam.com/docs/vbahv ... lume&ver=8
and that is my case.

VG RP space amount is now almost 6TB (the 2 VM only use 1,3TB): is possible to let Veeam give an expiration day to these VG Recovery Points?

Thank you.

PAP

[ronnmartin61]

@PAP400 I did a cursory check and I seem to be able to reproduce the behavior where snapshots are retained until default expiry like you mention. Would you mind going ahead and opening a case as I do not believe this is the expected behavior

[PAP400]

Hi ronnmartin61,

thank you for the reply.

Just opened case # 07827586 — Issue with Recovery Points / Snapshots on Nutanix cluster with Veeam.


PAP400

[ronnmartin61]

@PAP400 thanks! I'll track on this end

[PAP400]

Hello,

just to notify that the [Case #07827586 — Issue with Recovery Points / Snapshots on Nutanix cluster with Veeam] is still open but with no solution at the moment.

VBR (VBAHV v7) leaves the VG Snapshot expiration period undefined; in this case AOS forcibly marks an expiration period of 60 days; Nutanix's usual cleanup activities then delete VG snapshots older than 60 days.
IMHO, VBR should "mark" these VG Snapshots with a very low retention period and not with an undefined one.

My problem is that if a VM, that uses a VG as storage, daily changes many blocks, I will have 60 huge VG snapshots, with the result that the space used by VG snapshots is far more (600% at the moment) the space provisioned by the VM itself.
In my case this implies that now my cluster can not afford a node fail due to the significant space taken up by the 60 VG snapshots for each VG on the main Storage Pool.

Stay tuned

PAP400

[ronnmartin61]

@PAP400 definitely not the expected behavior. This has been escalated to engineering internally...