Cluster Operators degraded after migrating from OpenShift SDN to OVN-Kubernetes in RHOCP4

Issue

• Critical cluster operators degraded after migrating from OpenShift SDN CNI to OVN-Kubernetes.

Etcd

      message: 'ClusterMemberControllerDegraded: unhealthy members found during reconciling members

      EtcdEndpointsDegraded: EtcdEndpointsController can''t evaluate whether quorum
      is safe: etcd cluster has quorum of 2 and 2 healthy members which is not fault
      tolerant: [{Member:ID:1730761836105137485 name:"master-2.example.com"
      peerURLs:"https://10.186.22.11:2380" clientURLs:"https://10.186.22.11:2379"  Healthy:false
      ....

Authentication

      message: 'WellKnownAvailable: The well-known endpoint is not yet available: failed
      to GET kube-apiserver oauth endpoint https://10.186.22.11:6443/.well-known/oauth-authorization-server:
      dial tcp 10.186.22.11:6443: i/o timeout'

• The controlplane node generates events with warnings related to ErrorReconcilingNode caused due to failing SNAT sync for Node Feature Discovery (NFD) operator pods.

Events:
  Type     Reason                Age                       From          Message
  ----     ------                ----                      ----          -------
  Warning  ErrorReconcilingNode  3m27s (x3219 over 2d20h)  controlplane  error creating gateway for node master-2.example.com: failed to init shared interface gateway: failed to sync stale SNATs on node master-2.example.com: unable to fetch podIPs for pod openshift-nfd/nfd-master-aqmt