Developer console dashboards empty after Red Hat Advanced cluster manager import

Issue

Developer console dashboards are empty after the cluster is imported into Red Hat Advanced cluster manager.
This can also be caused by the manual addition of a Prometheus externalLabels setting a value for cluster=

Depending on the navigation path taken in the UI the query used may or may not contain the label cluster=

One example of this is the query for "CPU usage"
If the following navigation is used

Developer console > observe > dashboard - "CPU usage" Panel -> Inspect -> "show promql" the query is - 

sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{cluster="", namespace="$NAMESPACE"}) by (pod) 

But if the navigation used is:
Developer console > observe > metrics - "select query" - "CPU usage" - "show PromQL" the query is

(pod)sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{namespace='$NAMESPACE'}) by (pod)

Notice how the cluster= label is used in the first query but not the second