Document how to create dashboard & other key components

[justinsb]

No description provided.

[azell]

I installed a fresh k8s 1.3.4 cluster on AWS and find that the dashboard has been installed automatically. However the dashboard pod has errors:

Starting HTTP server on port 9090
Creating API server client for https://100.64.0.1:443
Error while initializing connection to Kubernetes apiserver. This most likely means that the cluster is misconfigured (e.g., it has invalid apiserver certificates or service accounts configuration) or the --apiserver-host param points to a server that does not exist. Reason: the server has asked for the client to provide credentials

I can hit the master over HTTPS though the /ui URL shows:

{
  "kind": "Status",
  "apiVersion": "v1",
  "metadata": {},
  "status": "Failure",
  "message": "no endpoints available for service \"kubernetes-dashboard\"",
  "reason": "ServiceUnavailable",
  "code": 503
}

Maybe a bad node?

[azell]

This error might disappear when I rm -fr $HOME/.terraform.d/. But daemonsets that were running on a destroyed cluster (newrelic-agent) were automatically running when I re-created the cluster. Is kops caching that info somewhere?

After running terraform to create the cluster:

kubectl --namespace=kube-system get po
NAME                                                   READY     STATUS    RESTARTS   AGE
etcd-server-events-ip-172-20-83-15.ec2.internal        1/1       Running   1          19m
etcd-server-ip-172-20-83-15.ec2.internal               1/1       Running   1          18m
kope-aws-ip-172-20-70-54.ec2.internal                  1/1       Running   0          9s
kope-aws-ip-172-20-83-15.ec2.internal                  1/1       Running   1          18m
kube-apiserver-ip-172-20-70-54.ec2.internal            1/1       Running   3          18s
kube-apiserver-ip-172-20-83-15.ec2.internal            1/1       Running   3          17m
kube-controller-manager-ip-172-20-83-15.ec2.internal   1/1       Running   1          18m
kube-dns-v14-rr0ca                                     2/3       Running   2          17m
kube-proxy-ip-172-20-83-15.ec2.internal                1/1       Running   1          18m
kube-proxy-ip-172-20-88-200.ec2.internal               1/1       Running   0          17m
kube-proxy-ip-172-20-88-201.ec2.internal               1/1       Running   0          17m
kube-proxy-ip-172-20-88-202.ec2.internal               1/1       Running   0          17m
kube-scheduler-ip-172-20-70-54.ec2.internal            1/1       Running   0          9s
kube-scheduler-ip-172-20-83-15.ec2.internal            1/1       Running   1          17m
newrelic-agent-1chhy                                   1/1       Running   0          18m
newrelic-agent-9jc8s                                   1/1       Running   0          18m
newrelic-agent-hlwvy                                   1/1       Running   0          18m
newrelic-agent-sql9y                                   1/1       Running   1          18m

I did not install the newrelic-agent daemonset though it did exist on a previous (destroyed) cluster with the same name and S3 bucket.

[justinsb]

I've written up how to install addons: https://github.com/kubernetes/kops/blob/master/docs/addons.md

@azell I think the problem is that the disks are persisting, I think because if you use the terraform workaround you need to kops delete cluster to delete the disks. With terraform 0.7 released, I'm going to see if we can get rid of the workaround, or document the need to kops delete cluster: #256

Going to close this issue in favor of #256. And if problems persist with the dashboard, please do open an issue for that also (or ping me directly!)

[azell]

Great, will use kops delete cluster.