Kubernetes

k3s Production Cluster on Linux with Traefik and Longhorn Basics

Author :Liam K.
Date :August 3, 2026
Time :22 minutes

k3s is a certified Kubernetes distribution optimized for edge and small production clusters. It ships with sensible defaults, uses less memory than full kubeadm stacks, and still supports Helm, Ingress, and CSI storage. This guide builds a multi-node cluster suitable for internal platforms and self-hosted apps.

Prerequisites

At least one control-plane VM and one worker (2+ CPU / 4 GB each)
Open ports between nodes: 6443, 10250, and flannel/VXLAN as required
DNS or MetalLB/cloud LB plan for Ingress

Step 1: Install the First Server Node

bash

curl -sfL https://get.k3s.io | sh -s - server \
  --write-kubeconfig-mode 644 \
  --tls-san k3s.example.com \
  --disable traefik=false
sudo kubectl get nodes
sudo cat /var/lib/rancher/k3s/server/node-token

Step 2: Join Worker Nodes

bash

# On each worker
curl -sfL https://get.k3s.io | K3S_URL=https://CONTROL_PLANE_IP:6443 \
  K3S_TOKEN='NODE_TOKEN_FROM_SERVER' sh -s - agent
# Back on server
sudo kubectl get nodes -o wide

Step 3: Configure kubectl Locally

bash

mkdir -p ~/.kube
sudo cp /etc/rancher/k3s/k3s.yaml ~/.kube/config
sudo chown $USER:$USER ~/.kube/config
sed -i 's/127.0.0.1/CONTROL_PLANE_IP/' ~/.kube/config
kubectl get pods -A

Step 4: Deploy a Sample App with Ingress

bash

kubectl create deployment whoami --image=traefik/whoami:v1.10
kubectl expose deployment whoami --port=80
kubectl apply -f - <<'EOF'
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: whoami
  annotations:
[...]

Command truncated. Copy to view full command.

Step 5: Add Longhorn for Persistent Volumes

bash

kubectl apply -f https://raw.githubusercontent.com/longhorn/longhorn/v1.7.2/deploy/longhorn.yaml
kubectl -n longhorn-system get pods
# Create a PVC using storageClassName: longhorn for stateful workloads

Step 6: Backup etcd and Plan Upgrades

bash

sudo k3s etcd-snapshot save --name pre-upgrade
sudo ls /var/lib/rancher/k3s/server/db/snapshots
# Upgrade control plane first, then agents, following k3s release notes
curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=v1.31.4+k3s1 sh -s - server

Production Checklist

Protect kubeconfig and the node token like root credentials.
Take etcd snapshots before upgrades and store them off-node.
Separate system and app node pools when the cluster grows.
Enable TLS on Ingress for public hostnames (cert-manager or Traefik ACME).
Monitor node disk pressure — Longhorn and container images fill disks quickly.

"Lightweight Kubernetes still needs real operations: snapshots, upgrades, ingress TLS, and storage capacity planning."

Technical Author

Liam K.

System administrator and technical writer specializing in server infrastructure, security and deployment. Creating comprehensive guides to help you master server administration.