When stopping the VMs making up a 4.19.x cluster and waiting more than the 24h window where certs get rotated after the install the cluster does no longer recover.

This had been (finally) fixed in 4.4.8. And it now reappears. This is a critical capability for customers (and the demo platform team) to save money when clusters are not in use.

    

    How reproducible:{code:none}

    

    1. Deploy a 4.19 cluster
    2. After the cluster has settled (all ClusterOperators happy etc.) stop the VMs making up the cluster.
    3. Wait 30 hours (to be sure to miss the 24h cert rotation window)
    4. Start the VMs.
    5. Wait for CSRs to appear. As Pending CSRs appear approve them.
    6. Nodes will report Ready, all ClusterOperators will report Ready
    7. OpenShift api services are not ready (oc get apiservices).
    

    Can't access Openshift Console. Can't execute openshift specific commands (oc project, ...
    

    Cluster recovers and is operational