Kubernetes 1.7 is there and our Vagrant kubeadm setup does not work anymore

[rmohr]

Yes it is that time again :)

Kubeadm tries to deploy Kubernets 1.7 after a fresh vagrant up. The master comes up properly, but the node can't register with the provided token.

The error message is:

==> node0: [discovery] Trying to connect to API Server "192.168.200.2:6443"
==> node0: [discovery] Created cluster-info discovery client, requesting info from "https://192.168.200.2:6443"
==> node0: [discovery] Failed to connect to API Server "192.168.200.2:6443": there is no JWS signed token in the cluster-info ConfigMap. This token id "abcdef" is invalid for this cluster, can't connect

There were some changes regarding to the token and bootstrapping.

[rmohr]

/cc @kubevirt/maintainers

[rmohr]

I can start working on this shortly, but if someone has time right now, feel free to pick it.

[rmohr]

Got it: kubernetes/kubeadm#335

[rmohr]

Provided a temporary fix in #288

[stu-gott]

As mentioned in kubernetes/kubernetes#48480 the root cause of this appears to be a race condition (already merged into upstream git). In order to get Kubernetes 1.7 working before that patch hits, I simply recreated my vagrant nodes until everything came up properly. It took two re-creates for me. YMMV

[vhosakot]

Per kubernetes/kubeadm#335 (comment), kubeadm init --token-ttl 0 resolves this issue for me.