VMHA - CEPH - Why is NFS pri storage needed?

[returntrip]

We are testing Linstor and CEPH as primary storage backends. They both seems to work fine and the VMs are being restarted after 5 minutes time failure.

Whilst Linstor does not need a dummy NFS pri storage, CEPH requires a dummy NFS storage as described here (in my case the NFS pri storage must be enabled tho). Can someone please explain why is this needed for VM HA tp work on CEPH pri storage?

[weizhouapache]

@returntrip
you are right. the NFS pool must be enabled prior to #11666
you can tag it.

for VM HA on ceph storage, there was a PR but not merged yet, refer to #5862

[returntrip]

Thanks for the reply!