I'm updating the bug title (and moving the test-case name into the environment field for Sippy) to make it more clear what is getting fixed here.

checked with CI jobs, no firing AlertmanagerMembersInconsistent alert
https://search.ci.openshift.org/?search=AlertmanagerMembersInconsistent&maxAge=48h&context=1&type=bug%2Bjunit&name=periodic-ci-openshift-release-master-ci-4.8.*&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

    - alert: AlertmanagerMembersInconsistent
      annotations:
        description: Alertmanager {{ $labels.namespace }}/{{ $labels.pod}} has only
          found {{ $value }} members of the {{$labels.job}} cluster.
        summary: A member of an Alertmanager cluster has not found all other cluster
          members.
      expr: |
        # Without max_over_time, failed scrapes could create false negatives, see
        # https://www.robustperception.io/alerting-on-gauges-in-prometheus-2-0 for details.
          max_over_time(alertmanager_cluster_members{job="alertmanager-main",namespace="openshift-monitoring"}[5m])
        < on (namespace,service) group_left
          count by (namespace,service) (max_over_time(alertmanager_cluster_members{job="alertmanager-main",namespace="openshift-monitoring"}[5m]))
      for: 15m
      labels:
        severity: critical

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (OpenShift Container Platform 4.8.5 security update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2021:3121