* Add a reboot-coordinator namespace for CLUO components * Define an RBAC ClusterRole for update-operator and update-agent * Replace the older-style where CLUO ran in kube-system, with admin privilege