Any update on this?
This looks the same as https://bugzilla.redhat.com/show_bug.cgi?id=1952819 to me. We can see in ovn-controller there was no binding for alertmanager-main-1. That indicates to me that the entry is not in the db. Can you confirm? Either attach the db or exec into the nbdb:
oc exec -n openshift-ovn-kubernetes ovnkube-master-4tljb -c nbdb ovn-nbctl --no-leader-only list logical_switch_port openshift-monitoring_alertmanager-main-1
If there is no row (entry), then restart the *active* ovnkube-master pod. Check and see if this instance becomes active again, if not you may need to restart other ovnkube-masters. This should effectively re-initialize the cache with the logical port missing and correctly create the missing port in the db. If this works then it is definitely 1952819.