Hi @Geoff Burke ,

I haven’t faced issues directly with K10 when used along with longhorn recently. You could take a look at this KB article https://kb.kasten.io/knowledge/debugging-backups-with-longhorn-csi

There was one small problem with the snapshot cleanup(It was not an issue with K10. Its just how longhorn snapshot garbage collection works) that I faced when I was working along with @Lei Wei .

I have drafted a KB article for that issue with the solution. Its not published yet. I will share the article once it is public. 

Thanks jaigneshjk! I will give it a try later today. 

cheers

@jaiganeshjk All good. The only thing that I had to do was create the volumesnapshot class (type snap) and worked like a charm. By the way in my new super duper lab   I am using both Longhorn and Rook Ceph. Is there a way to have Veeam automatically choose which one to use or can you have two or do I have to keep changing the annotation?

Glad to hear that it worked.

 I am using both Longhorn and Rook Ceph. Is there a way to have Veeam automatically choose which one to use or can you have two or do I have to keep changing the annotation?

You can have both drivers and their corresponding volumesnapshotclass.

K10 uses the volumesnapshotclass based on the PVC and the provisioner that provisioned the PVC.

You can specify default annotation for both longhorn and rook-ceph volumesnapshotclass as they both use different provisioners, and K10 will choose the correct volumesnapshotclass based on the PVC that is protected.

Excellent thanks that makes perfect sense now that I think about it 

I have drafted a KB article for that issue with the solution. Its not published yet. I will share the article once it is public. 

@Geoff Burke 

Below is the link to the KB article that I promised to share earlier.
https://kb.kasten.io/knowledge/too-many-snapshots-longhorn

Thanks @jaiganeshjk