OpenShift cluster fails to upgrade due to MCO in Degraded state.

Issue

• Cluster Upgrades fail intermittently with degraded MCO.
• The machine-config-controller pod shows the following logs:

1 template_controller.go:421] Error syncing controllerconfig machine-config-controller: failed to read dir "/etc/mcc/templates": readdirent /etc/mcc/templates: no such file or directory
1 render_controller.go:382] ControllerConfig has not completed: completed(false) running(false) failing(true)
1 render_controller.go:383] Dropping machineconfigpool "worker" out of the queue: ControllerConfig has not completed: completed(false) running(false) failing(true)
1 render_controller.go:382] ControllerConfig has not completed: completed(false) running(false) failing(true)