OpenShift Virtualization 4.18 on x86 and s390x

    VM is deployed based on ODF. It is able to Live Migrating between nodes. That means, in a non planned situation like Node failure or Network cut the VM running on the failed/cutoff node must be automatically failovered to an other node. It should happen the similar way like a [stateless | statefull] Deployment with ReplicaSet. However, it does not happen. The underlying pod gets into Terminating state on the failed node forever. And the VM stays in status Running on the poweroffed node. 

Problem: The possible critical workload on the VM becomes unavailable. RTO is not fulfilled.      

    1. OCP 4.18 on x86 or on s390x. ODF is installed. 
    2. Deploy VM based on ODF (openshift-csi-rdb-virtualization as a default storage class)
    3. Poweroff the node, the VM is running on. 
    

    VM is not failovered to the healthy node. Instead, it stays in status Running. The underlying pod gets in status Terminating forever. 

    Once the poweroffed node has being monitored as NotReady, the VM should failover to a healthy node. It is the basics of the HA business.

- When the failed node is powered on and Reday again, then the VM continue to run.  
- Manually the failed node can be drained and deleted from the cluster. In this case the VM will be started on the other node. 
BUT: HA requires, it should happen automatically in the seconds / minutes scope of time, without any admin intervention.