Cloud-Credential-Operator in degraded state with CredentialsFailing and DeleteConflict RHOCP 4

Issue

• cloud-credential-operator (CCO) failed to sync some of the CredentialsRequest and threw a DeleteConflict error:

  2022-06-07T07:00:12.540335561Z time="2022-06-07T07:00:12Z" level=error msg="DeleteConflict: Cannot delete entity, must remove users from group first.\n\tstatus code: 409, request id: 123x456-789y-0123z
  

• After rotating the access keys and secret keys stored in aws-creds in the kube-system namespace and deleting all the CredentialsRequests for them to pick up the new credentials there are some CredentialRequests, which are not getting synced with, for example (openshift-machine-api-aws, openshift-ingress,openshift-image-registry, etc..) are failing to sync, when checking the status of these it shows following deprovisioning error:

  - lastProbeTime: "2022-06-07T05:02:41Z"
    lastTransitionTime: "2022-06-07T05:02:41Z"
    message: 'failed to deprovision resource: AWS Error: DeleteConflict: Cannot delete
      entity, must remove users from group first., status code: 409'
    reason: CloudCredDeprovisionFailure
    status: "True"
    type: CredentialsDeprovisionFailure