After throwing up a Rancher 2.0 cluster I attempted to install the K8S metrics-server deployment, which constantly crashed citing “Could not create the API server: cluster doesn’t provide requestheader-client-ca-file”. After additional reviews, Heapster was attempted and it continually dumped errors messages such as:
k8s.io/heapster/metrics/heapster.go:319: Failed to list *api.Pod: pods is forbidden: User “system:serviceaccount:kube-system:default” cannot list pods at the cluster scope
k8s.io/heapster/metrics/sources/kubelet/kubelet.go:339: Failed to list *api.Node: nodes is forbidden: User “system:serviceaccount:kube-system:default” cannot list nodes at the cluster scope
k8s.io/heapster/metrics/processors/node_autoscaling_enricher.go:100: Failed to list *api.Node: nodes is forbidden: User “system:serviceaccount:kube-system:default” cannot list nodes at the cluster scope
k8s.io/heapster/metrics/heapster.go:327: Failed to list *api.Node: nodes is forbidden: User “system:serviceaccount:kube-system:default” cannot list nodes at the cluster scope
I’ve also attempted to use Prometheus and Grafana from the Rancher catalog, however, neither fulled came online and both hung in the ‘ContainerCreating’ phase.
As with vanilla K8S installing the metrics-server was sufficient to get the pod metrics and enable HPA functionality I’m hoping something similarly straightforward exists for Rancher and I just haven’t discovered it yet.
Any pointing in the right direction would be greatly appreciated.