Right now, the client that the ClusterClientStore caches lives even though a Cluster object is marked unschedulable. We need to return a different error when a Cluster object is marked unschedulable, so that the controllers can be aware of it and remove the relevant items off their queues.
Right now, the client that the ClusterClientStore caches lives even though a Cluster object is marked unschedulable. We need to return a different error when a Cluster object is marked unschedulable, so that the controllers can be aware of it and remove the relevant items off their queues.