Btw. even though one node seems to be running fine, the cluster is stuck in "updating" mode, so I cannot go into the Rancher kubectl of that cluster to delete the node from there.