The Operator sometimes creates Ingress resource on OpenShift clusters. This should not be happening as on OpenShift a Route should be created instead.

It sometimes takes a long time (around a week) before the issue manifests.

This is especially problematic if there's more than 1 Keycloak instance in the cluster as the default Ingress uses a hardcoded "keycloak.local" host which then causes conflicts.

This was reproduced on OCP 4.11 and 4.13.