Customers VMs were running on a VMware ESX host in one datacenter and moved them to a Nutanix ESX host located in a different data center, but still within the same VMware hypervisor. After this migration they started seeing issues where the storage operator was unable to identify the VM UUIDs and was treating restarted nodes as new nodes.

To address this RedHat recommended updating the new datacenter name in the configmap

Updated this in cloud-provider-config in openshift-config namespace
[Workspace]
server = "nprd-vcenter01.bbtnet.com"
datacenter = "W11-NTX-SERVER-HOSTING"
default-datastore = "W11-OCP-NPRD-NTX-CL03-CN01"
folder = "/W11-NTX-SERVER-HOSTING/vm/OpenShift/ncw-az1-001"

[VirtualCenter "nprd-vcenter01.bbtnet.com"]
datacenters = "W11-NTX-SERVER-HOSTING"

Merged 2 kubelet configs as below
spec:
kubeletConfig:
allowedUnsafeSysctls:

kernel.msg*
autoSizingReserved: true
machineConfigPoolSelector:
matchLabels:
pools.operator.machineconfiguration.openshift.io/worker: ''
(comment #30 from the case https://gss--c.vf.force.com/apex/Case_View?id=500Hn00001pE4Gp&sfdc.override=1#comment_a0aHn00000ZRURoIAP)
WHen these changes happened customers non production were in OCP 4.16 version where they see issues. Change sequence was

Non production cluster OCP 4.14 upgrade to OCP 4.16 to vmotion migration to Updating config map- (intermittent outbound connectivity issues in OVN)
Production clusters OCP 4.14 to vmotion migration changes to Updating config map to OCP 4.16-(There are no issues).

    

 Its not supported and when QE  tested cluster was degraded.  Customer has replicated the issue.

    

    1.
    2.
    3.