K3D recipes

k3d is a lightweight wrapper to run k3s (Rancher Lab’s minimal Kubernetes distribution) in docker.

This repo contains a set of recipes to configure and use k3d in different use cases.

K3D installation

Installation requirements

K3d requires to have installed the following components

• docker
• kubectl

Installation procedure

To install, executes the following command:

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

Creating cluster using a config file

Create config file containing configuration of the cluster

k3d-cluster.yml

apiVersion: k3d.io/v1alpha3
kind: Simple
name: mycluster-simple
servers: 1
agents: 2

This is a basic configuration creating a cluster with 1 k3s server and 2 k3s agents

And execute

k3d cluster create -c k3d-cluster.yml

K3D behind a corporate HTTP proxy

HTTP_PROXY and HTTPS_PROXY environment variables need to be passed to k3d containers

As stated in K3d FAQ, it also requires to validate corporate certificate within the containers. For doing that path to corparte CA certificate must be mounted in k3d containers.

apiVersion: k3d.io/v1alpha3
kind: Simple
name: mycluster-behind-corporate-proxy
servers: 1
agents: 2
volumes:
  - volume: /etc/ssl/certs/corporate-ca.pem:/etc/ssl/certs/corporate-ca.crt
env:
  - envVar: HTTP_PROXY=http://<corporate_proxy>:<port> 
    nodeFilters:
      - all
  - envVar: HTTPS_PROXY=http://<corporate_proxy>:<port>
    nodeFilters:
      - all

Enabling access to K3s metrics endpoints

k3s arguments need to be passed to enabling remote monitoring to k3s metrics endpoints. Within configuratoin file set the parameters under options.k3s.extraArgs

apiVersion: k3d.io/v1alpha3
kind: Simple
name: mycluster-with-metrics-exposed
servers: 1
agents: 2
options:
  k3s:
    extraArgs:
      - arg: --kube-proxy-arg=metrics-bind-address=0.0.0.0
        nodeFilters:
          - server:*
          - agent:*
      - arg: --kube-controller-manager-arg=bind-address=0.0.0.0
        nodeFilters:
          - server:*
      - arg: --kube-scheduler-arg=bind-address=0.0.0.0
        nodeFilters:
          - server:*

Making traefik ingress accesible

For making accesible in localhost at 0.0.0.0:8080 the k3s loadbalancer where traefik is listening create the following k3d config file

apiVersion: k3d.io/v1alpha3
kind: Simple
name: mycluster-with-traefik-ingress-accesible
servers: 1
agents: 2
ports:
  - port: 8080:80 # same as `--port '8080:80@loadbalancer'`
    nodeFilters:
      - loadbalancer