magnuswahlstrand/local-k3s

Setup VMs

Launch three nodes

multipass launch -n n1
multipass launch -n n2
multipass launch -n n3

Setup K3S

Initiate K3S control node (takes < 20s on my machine)

multipass exec n1 -- bash -c "curl -sfL https://get.k3s.io | sh -"

Get access token for joining, store IP

TOKEN=$(multipass exec n1 sudo cat /var/lib/rancher/k3s/server/node-token)
IP=$(multipass info n1 | grep IPv4 | awk '{print $2}')

Store k3s configuration to host

multipass exec n1 sudo cat /etc/rancher/k3s/k3s.yaml > k3s.yaml

In order to access the host, we need to change the IP from 127.0.0.1 to the IP of the VM (n1)

sed -i '' "s/127.0.0.1/$IP/" k3s.yaml

Finally, tell kubectl to use the k3s configuration file to talk to the configuration cluster

export KUBECONFIG=$PWD/k3s.yaml

multipass exec n2 -- bash -c "curl -sfL https://get.k3s.io | K3S_URL=\"https://$IP:6443\" K3S_TOKEN=\"$TOKEN\" sh -"
multipass exec n3 -- bash -c "curl -sfL https://get.k3s.io | K3S_URL=\"https://$IP:6443\" K3S_TOKEN=\"$TOKEN\" sh -"

Other

sudo kubectl patch serviceaccount default \
               -p '{"imagePullSecrets": [{"name": "gcr-json-key"}]}'

Verify that secret has been added

sudo kubectl get serviceaccount default -o yaml

Resources

• Using Google Container Registry with Kubernetes - Li Lian