Alertmanager Pod keeps recreating itself and never deploys in OpenShift 4

Issue

• Among multiple Alertmanager-main pods, alertmanager-main-1 keeps recreating itself and never deployed, while alertmanager-main-0 has deployed.

$ oc get sts -n openshift-monitoring alertmanager-main
NAME                READY   AGE
alertmanager-main   1/2     4d

$ oc get pods -n openshift-monitoring -l alertmanager=main
NAME                  READY   STATUS        RESTARTS   AGE
alertmanager-main-0   6/6     Running       0          4d
alertmanager-main-1   0/6     Terminating   0          0s

• Re-creating the alertmanager pod or Statefulset manually will not improve the situation.

• The event log shows a message that Statefulset is starting a rolling update of alertmanager-main-1, but the pod is terminated immediately after it is created.

$ omc -n openshift-kube-controller-manager logs kube-controller-manager-master-0.fd-sandbox.mano.local -c kube-controller-manager | tail -n 2
2022-XX-XX       1 stateful_set_control.go:571] "Pod of StatefulSet is terminating for update" statefulSet="openshift-monitoring/alertmanager-main" pod="openshift-monitoring/alertmanager-main-1"
2022-XX-XX       1 event.go:294] "Event occurred" object="openshift-monitoring/alertmanager-main" fieldPath="" kind="StatefulSet" apiVersion="apps/v1" type="Normal" reason="SuccessfulDelete" message="delete Pod alertmanager-main-1 in StatefulSet alertmanager-main successful"

• Also, Prometheus-k8s of statefulset has not been deployed.

$ oc get sts -n openshift-monitoring prometheus-k8s
No resources found in openshift-monitoring namespace.

• How do I get the Alertmanager and Prometheus-k8s pods to deploy and collect metrics successfully?