Travis Nielsen

In this case I'm thinking we can leave this issue active to track when the csi driver update is integrated to rook.

Just to be clear, which deployments have keyrings associated with them? mds? any others? I don't recall offhand.

The error "is still being used" seems to indicate that the mount isn't allowed as long as another pod has it mounted. Did you run "rbd unmap" directly? The CSI...

Yeah the rule was removed based on [this discussion](https://github.com/rook/rook/issues/10070#issuecomment-1099455108), and it does appear unintentional that it came back without being fixed. @jdloft Are you able to test that new condition...

> It seems that no quorom was established so far > > ``` > 2022-07-08 11:55:29.229357 E | op-mon: attempting to continue after failing to start mon "a". failed to...

Perfect! No rush on this feature.

I would think the INPUT_FILE is a temporary file that has the settings to be applied, then it's deleted after the assimilate-conf command is run.

As a test workaround, does the mon stay healthy if the [liveness probe is disabled](https://rook.io/docs/rook/latest/CRDs/Cluster/ceph-cluster-crd/#health-settings)? You could also try adjusting the timeouts or other parameters on the liveness probe.

@guimou Shall we reopen this PR to finish the documentation?

See the commitlint and DCO CI failures. To your commit message add the "docs:" prefix, and your commit message needs an email signature for the DCO.