Harrison

After further investigation I've found that the issue with `dashboard-shell` only reproduces on RKE2 clusters. On RKE1 clusters the `dashboard-shell` container will pull the `shell` image from the proper cluster...

It looks like this is due to the RKE2 cluster agent being given the wrong value for the `CATTLE_CLUSTER_REGISTRY`. The `dashboard-shell` container will pull the `shell` image from the correct...

/backport v2.6.10

sync'd up with gaktive and trasnferred this ticket to rancher/rancher, as it does not seem to be a UI issue. > I asked Seb about it - it sounds like...

Build failure will be addressed by https://github.com/rancher/rancher/pull/39347

#39347 has been merged, allowing this PR to pass CI. @tuxtof please rebase this PR on top of `release/v2.7`, the build will pass on the next try.

This issue seems to be related to https://github.com/rancher/rke2/issues/1053, both K3s and RKE2 require that the `nm-cloud-setup.service` and `nm-cloud-setup.timer` are disabled for provisioning to work correctly on RHEL 8.4, and it...

After looking into this a bit more I am confident that the root cause of this issue is due to the `nm-cloud-setup.service` and `nm-cloud-setup.timer` services not being disabled upon initial...

Machine version within rancher has been bumped to v0.15.0-rancher95, this issue should be OK for testing.

### Root Cause Rancher-machine changed how the hostname was set within the cloud-init when provisioning VSphere nodes. `local-hostname` is [the correct parameter for the NoCloud datasource.](https://cloudinit.readthedocs.io/en/latest/topics/datasources/nocloud.html) ### What was fixed...