Description of problem:

The below testcase are failing in upgrade jobs

1.  [sig-network] pods should successfully create sandboxes by other
2. [bz-monitoring][invariant] alert/Watchdog must have no gaps or changes
3. API LBs follow /readyz of kube-apiserver and stop sending requests before server shutdowns for external clients
4. [sig-network] pods should successfully create sandboxes by adding pod to network 
5. [sig-instrumentation] Prometheus [apigroup:image.openshift.io] when installed on the cluster shouldn't report any alerts in firing state apart from Watchdog and AlertmanagerReceiversNotConfigured [Early][apigroup:config.openshift.io] [Skipped:Disconnected] [Suite:openshift/conformance/parallel] 

Joblink - 1954878323259609088