Besides rebooting the cluster, is there anything you can do about a resource that is stuck "unloading".