Travis Nielsen

> reporting sidecar versions are also useful. Do we detect the sidecar versions currently? If so we can certainly report it. We just can't just report the image names directly...

Sounds good to rename these example labels, some others were already previously updated, but these were missed. @humblec Are you planning to open the PR?

@jmolmo Do you know what storage class the dashboard is looking for? Looks like we need a fix in the rook orchestrator module

Correct, `ceph version` only shows the version of the ceph client in the image. `ceph versions` will show what all the daemons are running. Since the mgr would be running...

@parth-gr Could you summarize the issue in the title and body instead of just adding the link? thanks

Setting `noout` and related flags has been discussed in the past around several scenarios such as during upgrades, or during node upgrades. The PDB management does currently set `noout` when...

> Thank you, @travisn. The proposed approach looks good to me. When all the related tests would pass with `skipSetOSDUpgradeFlags: true` then the default can be changed to `true`? I...

Per discussion with @BlaineEXE, an improvement on the proposal would be to only mark the specific OSD IDs as `noout` for which upgrade is actively in progress. If an OSD...

Agreed on those points to make sure the noout does not stay enabled, and we will need to coordinate between the OSD and PDB reconciles.

The ceph version inside the base image for the toolbox is the same as the operator, it might not match the running version of the ceph daemons. That said, there...