Last modified September 4, 2017

Adding Metrics with Heapster

Heapster is a simple monitoring solution that supports Kubernetes and CoreOS natively. Although we usually prefer Prometheus, currently you need Heapster for integrating metrics in Kubernetes Dashboard or running the Horizontal Pod Autoscaler.

You can deploy Heapster as a standalone solution (recommended if you are running Prometheus for monitoring anyway) or as a complete monitoring solution, e.g. together with InfluxDB and Grafana (currently not documented here).

Deploy Heapster (standalone)

Deploying Heapster is very straight forward.

kubectl apply --filename https://raw.githubusercontent.com/kubernetes/heapster/master/deploy/kube-config/rbac/heapster-rbac.yaml
kubectl apply --filename https://raw.githubusercontent.com/kubernetes/heapster/master/deploy/kube-config/standalone/heapster-controller.yaml

If you have the Kubernetes Dashboard installed, you should be able to see metrics being shown in the workload overview as well as on the pods after a reload. If they don’t show up directly, just wait a moment and try again.