The alert "etcdGRPCRequestsSlow" fires in CI

(Initially reported as BZ in https://bugzilla.redhat.com/show_bug.cgi?id=2072219)

Description of problem:
The alert was fired on build02 during upgrade from 4.10.6 to 4.10.8
https://coreos.slack.com/archives/CHY2E1BL4/p1649150672833739

Everything went back to normal shortly after.

I found nothing outstanding following the runbook of the alert.
https://github.com/openshift/runbooks/blob/master/alerts/cluster-etcd-operator/etcdGRPCRequestsSlow.md

My questions are:
1. Are slow etcd requests expected to happen during upgrade?
In any case, this is the must-gather.
https://coreos.slack.com/archives/CHY2E1BL4/p1649168331222259?thread_ts=1649150672.833739&cid=CHY2E1BL4

2. The condition of the alert has never last over 10m. Yet it was fired. Why?
https://coreos.slack.com/archives/CHY2E1BL4/p1649184857585639?thread_ts=1649150672.833739&cid=CHY2E1BL4

—

Reported in slack forum most recently: https://coreos.slack.com/archives/C027U68LP/p1662670368663159