Description of problem:
In this CI run we see that the job failed because the kube-storage-version-migrator did not become available again:

https://prow.ci.openshift.org/view/gs/origin-ci-test/logs/release-openshift-origin-installer-e2e-aws-disruptive-4.8/1412175886651756544

fail [github.com/openshift/origin/test/extended/dr/quorum_restore.go:296]: Jul  6 00:29:44.297: Some cluster operators never became available /kube-storage-version-migrator

but if we look at the clusteroperator status dump from the time the test failed, we see that the "available" condition is set to false, while the message is "all is well".  Those two things seem incompatible w/ each other:


Jul  6 00:29:44.297: INFO: ClusterOperators:
NAMESPACE NAME                     PROGRESSING AVAILABLE VERSION MESSAGE
          authentication           False       True              All is well
          baremetal                False       True              
          cloud-controller-manager False       True              
          cloud-credential         False       True              
          cluster-autoscaler       False       True              
          config-operator          False       True              All is well
          console                  False       True              All is well
          csi-snapshot-controller  False       True              All is well
          dns                      False       True              
          etcd                     False       True              NodeControllerDegraded: All master nodes are ready
EtcdMembersDegraded: No unhealthy members found
 image-registry                           False True   
 ingress                                  False True   The "default" ingress controller reports Degraded=False.
 insights                                 False True   
 kube-apiserver                           False True   NodeControllerDegraded: All master nodes are ready
 kube-controller-manager                  False True   NodeControllerDegraded: All master nodes are ready
 kube-scheduler                           False True   NodeControllerDegraded: All master nodes are ready
 kube-storage-version-migrator            False False  All is well
 machine-api                              False True   
 machine-approver                         False True   
 machine-config                           False True   
 marketplace                              False True   Available release version: 4.8.0-0.ci-2021-06-14-090014
 monitoring                               False True   
 network                                  False True   
 node-tuning                              False True   DaemonSet "tuned" available
 openshift-apiserver                      False True   All is well
 openshift-controller-manager             False True   All is well
 openshift-samples                        False True   
 operator-lifecycle-manager               False True   
 operator-lifecycle-manager-catalog       False True   
 operator-lifecycle-manager-packageserver False True   
 service-ca                               False True   All is well
 storage                                  False True   AWSEBSCSIDriverOperatorCRDegraded: All is well




Version-Release number of selected component (if applicable):
4.8.0-0.ci-2021-06-14-090014

How reproducible:
unknown

Actual results:
operator is unavailable but message implies everything is fine


Expected results:
when operators are unavailable, their message should indicate why not