k3d is a lightweight wrapper to run k3s (minimal Kubernetes distribution from Rancher Labs) in Docker.

By using k3d, we can create single and multi node k3s clusters in Docker for local development purpose.

In this blog, we are going to use Ubuntu 20.04 LTS for setting up k3d.

---

Install Docker using the convenience script

$ curl -fsSL https://get.docker.com -o get-docker.sh

$ sudo sh get-docker.sh

Add your user to the docker group to execute commands as non-root user and log out from the current session

$ sudo usermod -aG docker $USER

Log back in and verify the Docker commands as non-root user

$ docker info

$ docker container run hello-world

Download and install k3d

$ curl -s https://raw.githubusercontent.com/k3d-io/k3d/main/install.sh | bash

$ k3d version

Download and install kubectl

$ curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl"

$ sudo install kubectl /usr/local/bin/kubectl

$ kubectl version --client

Create the cluster using the below configuration file

$ cat config.yml

apiVersion: k3d.io/v1alpha4
kind: Simple
metadata:
  name: dev
servers: 1
agents: 2
image: rancher/k3s:v1.25.2-k3s1
ports:
- port: 30000-30100:30000-30100
  nodeFilters:
  - server:*
registries:
  create:
    name: dev
    host: 0.0.0.0
    hostPort: "5000"
options:
  k3s:
    extraArgs:
    - arg: --disable=traefik
      nodeFilters:
      - server:*

$ k3d cluster create --config config.yml

Check the cluster details and nodes

$ kubectl cluster-info

$ kubectl get nodes

---

Create a sample index.html file

$ echo "Hello from Pod!" > index.html

Build a custom Docker image using the below Dockerfile

$ cat Dockerfile

FROM nginx:1.23
COPY index.html /usr/local/nginx/html

$ docker image build -t app:v1 .

Tag the custom built image and push to the container registry created by k3d

$ docker image tag app:v1 localhost:5000/app:v1

$ docker image push localhost:5000/app:v1

Deploy a pod using the custom built image

$ cat app-pod.yml

apiVersion: v1
kind: Pod
metadata:
  labels:
    run: app
  name: app
spec:
  containers:
  - image: dev:5000/app:v1
    name: app
    ports:
    - containerPort: 80

$ kubectl create -f app-pod.yml

$ kubectl get pods

Expose the pod to a node port service

$ cat app-nodeport.yml

apiVersion: v1
kind: Service
metadata:
  labels:
    run: app
  name: app-nodeport
spec:
  ports:
  - port: 80
    targetPort: 80
    nodePort: 30000
  selector:
    run: app
  type: NodePort

$ kubectl create -f app-nodeport.yml

$ kubectl get svc

Verify the webpage using a browser or curl

$ curl http://localhost:30000