After upgrading OpenShift to 4.5.16+ the cluster became unstable

Issue

• Getting rid of possible API Priority and Fairness Alpha issues, the cluster was upgraded to version 4.5.16+ as documented here, however the API continues to misbehave.
• The cluster was upgraded to version 4.5.16+ but the API is slow and some nodes become NotReady continuously.
• The prometheus-operator pod within openshift-monitoring presents strconv.ParseUint: parsing errors similar to the following ones:

2020-11-16T12:22:57.463660093Z E1116 12:22:57.463465       1 reflector.go:178] github.com/coreos/prometheus-operator/pkg/prometheus/operator.go:486: Failed to list *v1.PodMonitor: resourceVersion: Invalid value: "6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729": strconv.ParseUint: parsing "6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729/6443729": invalid syntax

NOTE: the bug was tracked via BZ#1891815 and the fix was released via 4.5.22 errata.