Kubernetes K3s installation on VPS
Home › Kubernetes on VPS: Install K3s in 15 Minutes

Kubernetes on VPS: Install K3s in 15 Minutes

June 2026 · 10 min read

Kubernetes is a powerful container orchestration platform, but full installation requires significant resources. K3s is a lightweight Kubernetes distribution developed by Rancher that uses only ~500MB of RAM and runs on a single VPS. It supports real workloads with full kubectl commands and can be upgraded to full-scale Kubernetes in the future.

System Requirements

Installing K3s

# Install K3s as single-node (master + worker on same machine)
curl -sfL https://get.k3s.io | sh -

# Check status
systemctl status k3s

# View nodes
kubectl get nodes
# Expected output
NAME     STATUS   ROLES                  AGE   VERSION
vps-01   Ready    control-plane,master   1m    v1.29.x+k3s1
kubeconfig: K3s stores the config at /etc/rancher/k3s/k3s.yaml — copy it to ~/.kube/config if you want to run kubectl as a regular user: mkdir -p ~/.kube && cp /etc/rancher/k3s/k3s.yaml ~/.kube/config && chown $(id -u):$(id -g) ~/.kube/config

Kubernetes Core Concepts

Deploy Your First App: Nginx

# Create a deployment
kubectl create deployment nginx --image=nginx:alpine --replicas=2

# Expose with a service
kubectl expose deployment nginx --port=80 --type=ClusterIP

# View status
kubectl get pods
kubectl get deployments
kubectl get services
# Or use a YAML manifest (recommended for production)
nano nginx-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx
  namespace: default
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:alpine
        ports:
        - containerPort: 80
        resources:
          requests:
            memory: "64Mi"
            cpu: "50m"
          limits:
            memory: "128Mi"
            cpu: "200m"
---
apiVersion: v1
kind: Service
metadata:
  name: nginx-svc
spec:
  selector:
    app: nginx
  ports:
  - port: 80
    targetPort: 80
kubectl apply -f nginx-deployment.yaml

Configuring Ingress Controller (Traefik)

K3s comes with the Traefik ingress controller built-in. Create an Ingress resource to route your domain:

nano nginx-ingress.yaml
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: nginx-ingress
  annotations:
    traefik.ingress.kubernetes.io/router.entrypoints: web
spec:
  rules:
  - host: myapp.example.com
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          service:
            name: nginx-svc
            port:
              number: 80
kubectl apply -f nginx-ingress.yaml

# Test (DNS must point to VPS IP first)
curl -H "Host: myapp.example.com" http://YOUR_VPS_IP

Automatic SSL with cert-manager

# Install cert-manager
kubectl apply -f https://github.com/cert-manager/cert-manager/releases/latest/download/cert-manager.yaml

# Wait until ready
kubectl wait --for=condition=ready pod -l app=cert-manager -n cert-manager --timeout=120s

# Create a ClusterIssuer for Let's Encrypt
cat <

Common kubectl Commands

# View all resources
kubectl get all

# View pod logs
kubectl logs POD_NAME -f

# Execute into a container
kubectl exec -it POD_NAME -- sh

# Scale replicas
kubectl scale deployment nginx --replicas=3

# Delete resources
kubectl delete deployment nginx
kubectl delete -f nginx-deployment.yaml

# View details
kubectl describe pod POD_NAME
kubectl describe node vps-01

Monitor Resource Usage

# Check metrics-server (already included in K3s)
kubectl top nodes
kubectl top pods

K3s makes Kubernetes accessible for standard VPS deployments — ideal for small to medium microservices or teams learning Kubernetes before scaling to full production clusters in the future.

Verify K3s is Working Correctly

# Check node status
kubectl get nodes
# Should show: STATUS = Ready

# Check all pods in the namespace
kubectl get pods --all-namespaces

# Check services are running
kubectl get services

# Test cluster DNS
kubectl run test-dns --image=busybox --rm -it --restart=Never -- nslookup kubernetes
Tip: If kubectl doesn't work, set KUBECONFIG: export KUBECONFIG=/etc/rancher/k3s/k3s.yaml or copy it to ~/.kube/config to use without sudo as a regular user

Common Issues & Troubleshooting

Error: "the server doesn't have a resource type" after installing K3s

K3s is not ready yet, or kubectl is not connected to the correct cluster:

# Check if K3s is running
systemctl status k3s

# Set kubeconfig
export KUBECONFIG=/etc/rancher/k3s/k3s.yaml
kubectl cluster-info

# View K3s logs if service won't start
journalctl -u k3s -n 50

Error: Pod stuck at "Pending" and not becoming Running

Usually caused by insufficient resources or node not Ready:

# View problem details
kubectl describe pod POD_NAME
# Check the Events section at the bottom

# Check node resources
kubectl describe node
# View Capacity vs Allocatable

# Verify node is Ready
kubectl get nodes -o wide

Error: Pod stuck in "CrashLoopBackOff"

Container crashed and keeps restarting. Check the logs to see what's wrong:

# View pod logs
kubectl logs POD_NAME
kubectl logs POD_NAME --previous   # logs from previous run

# Check container exit code
kubectl describe pod POD_NAME | grep -A3 "Last State"

Error: Ingress configured but not routing traffic

Verify the Traefik ingress controller is running and DNS points correctly:

# Check ingress resource
kubectl get ingress

# Check Traefik pod
kubectl get pods -n kube-system | grep traefik

# Test manually (DNS may not have propagated)
curl -H "Host: myapp.example.com" http://YOUR_VPS_IP

Error: K3s using too much RAM

K3s runs etcd, API server, and controller manager all on one machine. On 1GB VPS this may be tight:

# View K3s component resource usage
kubectl top pods -n kube-system

# On low RAM, reduce system pod resources
# Or use containerd instead of Docker: K3s uses containerd by default

# Check node resources
kubectl describe node | grep -A5 "Allocated resources"

Frequently Asked Questions

How is K3s different from full Kubernetes?

K3s is a slimmed-down Kubernetes distribution by Rancher for resource-constrained environments. It removes unnecessary cloud provider plugins, uses SQLite instead of etcd (or embedded etcd), and packages components into a single binary. kubectl commands work the same way, and you can upgrade to full-scale K8s clusters in the future.

Is a single VPS suitable for K3s?

Yes, for development, staging, small microservices, or non-critical workloads. 2GB+ RAM is recommended for production with multiple apps. The limitation is no failover — if the VPS goes down, all pods stop. For critical production, use a multi-node cluster.

How do I backup K3s cluster state?

K3s stores cluster state in SQLite at /var/lib/rancher/k3s/server/db/. Backup with cp -r /var/lib/rancher/k3s/server/db/ /backup/k3s-db-$(date +%Y%m%d) and backup all YAML manifests you used for deployments, since manifests are the source of truth for restores.

Related Articles