Single node

In this part, you will create a VM and run a single-node k0s on it.

Create the virtual machine

First, run the following command to launch a VM named k0s:

multipass launch -n k0s-1 --disk 10G --cpus 2 --memory 2G

Next, run a shell on that VM.

This will use the ubuntu user

multipass shell k0s-1

Create the cluster

First, from this shell, get the latest release of k0s.

curl -sSf https://get.k0s.sh | sudo sh

After a few tens of seconds, the k0s binary will be available in /usr/local/bin. Check the current version.

k0s version

This command returns a version like v1.33.4+k0s.0 (your version could be slightly different though, depending on when you’re following this workshop).

Next, use the following command to create a systemd unit file.

sudo k0s install controller --single

When installing k0s we can provide the following options:

--single allows to configure the instance as both a control plane and a worker Node
--worker allows to configure the instance as a worker Node

If none of these option is provided, the instance will act as a control plane only Node.

Next, start the cluster.

sudo k0s start

Then, after a few tens of seconds, verify the cluster is running properly.

sudo k0s status

You should get an output similar the following one.

Version: v1.33.4+k0s.0
Process ID: 1817
Role: controller
Workloads: true
SingleNode: true
Kube-api probing successful: true
Kube-api probing last error:

It takes a few tens of seconds for the cluster to be up and running. After this slight delay, you can verify that the k0s controller is started in systemd.

$ sudo systemctl status k0scontroller
● k0scontroller.service - k0s - Zero Friction Kubernetes
     Loaded: loaded (/etc/systemd/system/k0scontroller.service; enabled; preset: enabled)
     Active: active (running) since Tue 2025-09-16 23:08:10 CEST; 1min 58s ago
       Docs: https://docs.k0sproject.io
   Main PID: 1817 (k0s)
      Tasks: 108
     Memory: 583.4M (peak: 711.4M)
        CPU: 16.004s
     CGroup: /system.slice/k0scontroller.service
     ...

Communication from an external machine

As k0s comes with its own kubectl subcommand, you can communicate with the API Server directly from the Node (the k0s-1 VM you have created):

$ sudo k0s kubectl get node
NAME   STATUS   ROLES           AGE   VERSION
k0s-1  Ready    control-plane   11m   v1.33.4+k0s

You usually should not SSH into a cluster’s nodes to run kubectl commands. Instead, run these commands from an admin machine. In order to do so, run the following command from your local machine. It retrieves the kubeconfig file generated during the cluster creation (/var/lib/k0s/pki/admin.conf) and saves it in a local file named k0s.kubeconfig.

multipass exec k0s-1 -- sudo cat /var/lib/k0s/pki/admin.conf > k0s.kubeconfig

Next, replace localhost with the actual IP address of the Multipass VM.

K0S1_IP=$(multipass info k0s-1 | grep IP | awk '{print $2}')
sed -i'' "s/localhost/$K0S1_IP/" k0s.kubeconfig

Then, configure your local kubectl, so it uses this modified kubeconfig file.

export KUBECONFIG=$PWD/k0s.kubeconfig

You can now communicate with the newly created cluster from your machine. List the Nodes to make sure this is working properly.

$ kubectl get no
NAME   STATUS   ROLES           AGE   VERSION
k0s-1  Ready    control-plane   16m   v1.33.4+k0s

Testing the whole thing

Let’s now run a Deployment based on the nginx image and expose it though a NodePort Service (which makes the service accessible from outside the cluster using a specific port).

cat<<EOF | kubectl apply -f -
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx
spec:
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.26
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: nginx
spec:
  selector:
    app: nginx
  type: NodePort
  ports:
  - port: 80
    targetPort: 80
    nodePort: 30000
EOF

Make sure the resources have been created correctly.

kubectl get deploy,po,svc

You should get an output similar to the following one.

NAME                    READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/nginx   1/1     1            1           32s

NAME                         READY   STATUS    RESTARTS   AGE
pod/nginx-65dd7588fc-4nmwr   1/1     Running   0          32s

NAME                 TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
service/kubernetes   ClusterIP   10.96.0.1        <none>        443/TCP        6m56s
service/nginx        NodePort    10.110.235.193   <none>        80:30000/TCP   6s

Using k0s-1 IP address (192.168.64.22 in this example) and the NodePort 30000, we can access nginx.

nginx

In the next part, we’ll modify the default configuration.

Configuration →