3

I have installed the kubernetes cluster with https://github.com/kubernetes-sigs/kubespray with vagrant. The installation was successful.

Next I would like to call the dashboard and generate the dashboard token as follows: 

kubectl -n kube-system describe secret kubernetes-dashboard-token | grep 'token:' | grep -o '[^ ]\+$'

I've got the token and with kubectl proxy statement, it starts the dashboard service.

Then calling the url http://localhost:8001/api/v1/namespaces/kube-system/services/https:kubernetes-dashboard:/proxy/#!/login and enter the token, after then I've got:

enter image description here

What am I doing wrong?

This should be the solution: https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/troubleshooting-kubeadm/#default-nic-when-using-flannel-as-the-pod-network-in-vagrant

What do not know, how to solve.

softshipper
  • 32,463
  • 51
  • 192
  • 400
  • Have you tried checking if [flannel is working](https://github.com/kubernetes-sigs/kubespray/blob/master/docs/flannel.md)? You can also check `kubectl get all --all-namespaces` to see if all pods are up. – Crou Nov 18 '19 at 10:53
  • I am facing the same problem... I am using calico as this is default in kubespray. If I do `kubectl get all --all-namespaces` I see all resources in a proper RUNNING state – Jürgen Zornig Dec 02 '19 at 14:44
  • Have you managed to make it work using the info from comments and the answer? – Wytrzymały Wiktor Dec 11 '19 at 10:17
  • @softshipper Have you managed to find solution ? – Malgorzata Mar 19 '21 at 10:49

1 Answers1

1

I faced the same problem!

  1. Check if the proxy is started: kubectl proxy
  2. Update the version DashBoard: from v1.10.0 to v2.0.0-beta4
  3. Check if the NodePort is in the same dashboard's namespace
Samuel Santana
  • 11
  • 3