Hi team! I'm having problems with users not being able to use rancher-generated kubeconfigs since they seem to point to an old control plane IP address (which was changed after GKE credential rotation). How do I reconfigure it properly? people can do actions in rancher kubeshell and UI but kubeconfigs don't work... [UPD] giving more info: 1. we have 2 clusters for which we did the credential rotation. the cluster in which the rancher is deployed is fine, but the second one is having problems. 2. after the rotation, the second cluster was errored in rancher UI, but after some tweaks (basically after deleting the

cattle-token

and running the

rollout restart deployment cattle-cluster-agent

it appeared as "green" and seem to operate in the cluster just fine. 3. the problem occurs when users managed by Rancher + AzureAD are trying to get their new kubeconfig for the second cluster. as i said, it seem to be pointing to the old control plane IP. 4. the other problem which we discovered just today is that in the second cluster, when trying to execute shell, or watch pods logs, it appears as Disconnected.