Kubernetes 1.10.5 is unusable, auth is not working #43395
Description
Issue description
Kubernetes, as installed from nixos-unstable, is not usable, kube-dns is in a crash loop and I can't see logs or exec into any container. Helm does not work either because it fails to forward the port.
I have no KUBECONFIG environment variable, nor do I have a ~/.kube/config.
Steps to reproduce
in /etc/nixos/configuration.nix
services.kubernetes = {
# I have a SWAP device enabled. The following flag disables the error
# emitted by kubelet when a swap device is found.
# After all, this is a laptop machine and the cluster is for dev-only.
kubelet.extraOpts = "--fail-swap-on=false";
roles = [ "master" "node" ];
};and sudo -i nixos-rebuild test
$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
cratos Ready <none> 7m v1.10.5
$ kubectl get pods --all-namespaces
NAMESPACE NAME READY STATUS RESTARTS AGE
kube-system kube-dns-5746ddc44f-kvsrb 1/3 CrashLoopBackOff 9 7m
$ kubectl logs kube-dns-5746ddc44f-kvsrb -n kube-system -c kubedns
error: You must be logged in to the server (the server has asked for the client to provide credentials ( pods/log kube-dns-5746ddc44f-kvsrb))
$ nix-shell -p kubernetes-helm --run 'helm init'
$HELM_HOME has been configured at /home/kalbasit/.helm.
Tiller (the Helm server-side component) has been installed into your Kubernetes Cluster.
Error: forwarding ports: error upgrading connection: unable to upgrade connection: Unauthorized.What I tried
I have tried to add the admin user to system:masters by the following setting:
services.kubernetes = {
# I have a SWAP device enabled. The following flag disables the error
# emitted by kubelet when a swap device is found.
# After all, this is a laptop machine and the cluster is for dev-only.
kubelet.extraOpts = "--fail-swap-on=false";
roles = [ "master" "node" ];
apiserver.basicAuthFile = pkgs.writeText "users" ''
kubernetes,admin,0,"system:masters"
'';
};and add the following contents to ~/.kube/config
apiVersion: v1
clusters:
- cluster:
insecure-skip-tls-verify: true
server: https://127.0.0.1
name: local
contexts:
- context:
cluster: local
user: local
name: local
current-context: local
kind: Config
preferences: {}
users:
- name: local
user:
password: kubernetes
username: adminBut I'm still getting the same error:
$ kube logs kube-dns-5746ddc44f-kvsrb -n kube-system -c dnsmasq
error: You must be logged in to the server (the server has asked for the client to provide credentials ( pods/log kube-dns-5746ddc44f-kvsrb))Technical details
Please run nix-shell -p nix-info --run "nix-info -m" and paste the
results.
- system:
"x86_64-linux" - host os:
Linux 4.17.4, NixOS, 18.09pre145679.dae9cf6106d (Jellyfish) - multi-user?:
yes - sandbox:
no - version:
nix-env (Nix) 2.0.4 - channels(root):
"nixos-18.09pre145679.dae9cf6106d" - channels(kalbasit):
"nixos-18.09pre145524.2a8a5533d18" - nixpkgs:
/nix/var/nix/profiles/per-user/root/channels/nixos
relates to: