Travis Nielsen

If you specify `mgr.count: 1` in the cluster CR, this behavior of updating services would not be enabled. When there is only a single mgr, the services are not updated....

#10717 will help with this failover condition as long as the `defer` failover method is allowed to run to cleanup the aborted mon failover. This should be the case in...

Ok, good to see your confirmation that by removing the duplicate entry from the ceph.config, the operator was able to recover. Avoiding the duplicate entries would be a small fix...

I believe all the needed API is already there to support VFS in Rook, most of the configuration will be in the application pod instead of in Rook or the...

> Maybe it's more a documentation change than anything! Yes, it would be great to document these steps and create an example application yaml after you confirm it's working.

> considering the File mode has been GAd in 1.23 kubernetes version Are there any concerns with File mode before it GA'd? Just wanting to confirm if it's really a...

@humblec Do we need https://github.com/ceph/ceph-csi/pull/3204 in a csi release before we merge this?

> @travisn jfyi , finally its merged.. Ok, when is the next ceph-csi release planned? I assume we need to wait for that before merging this?

Per discussion with csi team, we will plan on keeping the csi version label even in rook master, and we can add a daily CI test against the csi canary...

> > Per discussion with csi team, we will plan on keeping the csi version label even in rook master, and we can add a daily CI test against the...