Dashboard: CPU/Memory Usage metrics not shown

[mdi...@gmail.com]

I am running a Kubernetes cluster of one master and two minions on CentOS VMs which was installed using the CentOS installation guide from kubernetes.io.

I then installed the dashboard using the yaml from the git repo (using the rawgit link).

The dashboard works, however I do not see the CPU/Memory Usage metrics on the "Workloads" and the "Conditions" section on the "Nodes" page.

Are these features only available on GCE instances? Or am I missing something?

[mdi...@gmail.com]

Replying to myself: The graphs don't come out of the box, I need to install Heapster with an InfluxDB backend and a Grafana UI.

This is described in this document: https://github.com/kubernetes/heapster/blob/master/docs/influxdb.md

For my hand-made installation also this document describing how to set up Heapster was useful: https://github.com/kubernetes/heapster/blob/master/docs/source-configuration.md

Currently the heapster image specified in deploy/kube-config/influxdb/heapster-deployment.yaml is wrong:

s|gcr.io/google_containers/heapster-amd64:v1.3.0-beta.1|gcr.io/google_containers/heapster-amd64:v1.3.0-beta.2|

Also I needed to edit deploy/kube-config/influxdb/grafana-service.yaml and activate "type: NodePort" since I don't use a real load balancer.