Summary:

 Scale down the telemetry-operator, so that it doesn't get in the way and try to manually configure Prometheus and the dashboards. So basically do something like this:

Goal:

• deploy kube-rbac-proxy in the Prometheus container
• Play with the kube-rbac-proxy, get familiar with it. Try curling Prometheus through the kube-rbac-proxy from the openstackclient container (I think we have some article on our confluence about how to include the kubernetes tokens in curl queries. I'll try to find it)
  1. Without including the kubernetes token, the curl should fail
  2. When you include the token, you should be able to get the metrics
• Try to manually configure the dashboard datasource
  1. Try creating a datasource, which doesn't go through the kube-rbac-proxy. You should be able to see somewhere that the dashboards are able to access Prometheus
  2. Try modifying the datasource to go through the kube-rbac-proxy. Is there some config option for specifying the token?