Follow-up from  https://issues.redhat.com/browse/OCPBUGSM-47749

If externalLabels is set like following on cluster-monitoring-config config-map

 prometheusK8s:
  externalLabels:
    cluster: application-cluster-name

 
and if any other metric from user workload/platform monitoring contains same label it will be replaced with whats been set on externalLabels section. You will see this change when you query metric from Thanos Querier. If you query same metric from Prometheus you will not see this change.