Now I have a Kubernetes cluster it is time to install Prometheus. I have worked with the Prometheus Operator however since my cluster only covers a single computer at this time I a bit concerned it will run unhappily. No time like the present to try though!

Time to use Helm to install it. Just to make life hard I’m going to use a beta release of Helm 3. Looks like we must start with adding repositories, which is great.

helm repo add stable https://kubernetes-charts.storage.googleapis.com/
helm install platform-prometheus stable/prometheus-operator
Error: apiVersion "monitoring.coreos.com/v1" in prometheus-operator/templates/exporters/kube-scheduler/servicemonitor.yaml is not available

Hmm, this yielded an unexpected error. Time to track down what I did not configure correctly! Sounds like a race condition of kind? Rerunning the command repeatedly definitely generates a new custom resource name on every run. I think this might have to do with my kube-apiserver running really slow for some reason.

That is a sad roadblock I will need to deal with later.