Skip to content

Development Setup

Exploring and trying out things hands-on is crucial to understand how things work. Thanks to modern development workflows running a Kubernetes cluster even on small computers is easily doable. This chapter focuses on setting up a local Kubernetes cluster on your Laptop.

A few tools will be mentioned. The possibilities are endless. In case you are already experienced, feel free to diverge from the recommendations. In case you want to be sure to follow allong, stick to the recommendations.

During the installation a few questions may come up. Write them down to ask them later. They should be answered in the "basics" section. The workshop will start with the development setup, so that you can already do some hands on in the basics chapter.

It is recommended to go with Linux or MacOs, but Windows should also be fine. Though, Windows is not tested thoroughly.

The cluster

The following section describes the setup of a local Kubernetes cluster. This cluster will be used for learning and exploring. The actual setup for the use within a home or remote server will be explained in the chapter "homeserver".

This workshop will show how minikube can be used to start a local Kubernetes cluster. Go to the URL below and follow install instructions:

Alternatives would be:

Setting up the runtime

The runtime is what minikube uses to spin up the Kubernetes cluster. Options vary from operating system to operating system. Note that some recommended options have a built in Kubernetes functionality which will not be used as part of this tutorial.

Windows

The easiest way with Windows is to install Docker Desktop and then use the "docker" driver of minikube. Alternatives are Rancher Desktop also with the "docker" driver, or installing Virtualbox using the "virtualbox" driver.

Note that using Docker Desktop for professional use requires licencing.

Linux

Linux has a wider variety of options. The easiest is still with installing either Docker Desktop or the Docker Engine which does not have any licencing requirements, even for professional use. Alternatives are Rancher Desktop also with the "docker" driver, or installing Virtualbox using the "virtualbox" driver.

Another option would be the installation of podman with using the respective driver.

MacOS

MacOS basically comes with the same options as Linux. The easiest is still with installing either Docker Desktop or the Docker Engine which does not have any licencing requirements, even for professional use. Alternatives are Rancher Desktop also with the "docker" driver, or installing Virtualbox using the "virtualbox" driver.

Another alternative for x86_64 Macs is the use of hyperkit.

Setting up the cluster

Minikube supports various "drivers" where the cluster can be started. Usually the default option is the best suitable one.

Setting up the cluster may look as follows:

❯ minikube start
😄  minikube v1.32.0 on Darwin 14.3.1 (arm64)
    ▪ KUBECONFIG=/Users/uib53682/.kube/config
✨  Automatically selected the docker driver
📌  Using Docker Desktop driver with root privileges
👍  Starting control plane node minikube in cluster minikube
🚜  Pulling base image ...
💾  Downloading Kubernetes v1.28.3 preload ...
    > preloaded-images-k8s-v18-v1...:  341.16 MiB / 341.16 MiB  100.00% 25.42 M
    > gcr.io/k8s-minikube/kicbase...:  410.58 MiB / 410.58 MiB  100.00% 15.58 M
🔥  Creating docker container (CPUs=2, Memory=7887MB) ...
🐳  Preparing Kubernetes v1.28.3 on Docker 24.0.7 ...
    ▪ Generating certificates and keys ...
    ▪ Booting up control plane ...
    ▪ Configuring RBAC rules ...
🔗  Configuring bridge CNI (Container Networking Interface) ...
    ▪ Using image gcr.io/k8s-minikube/storage-provisioner:v5
🔎  Verifying Kubernetes components...
🌟  Enabled addons: storage-provisioner, default-storageclass
🏄  Done! kubectl is now configured to use "minikube" cluster and "default" namespace by default

Minikube start will autoselect the best driver, but also accepts the option --driver='' to manually select the driver as discusseds in the previous section.

Minikube also had addons which make additional functionalities easily available. The addons can be installed directly with minikube start with adding the addons as an option in a comma separated list like --addons='ingress,metrics-server'. Or after the cluster start as follows.

# ingress is required later on
minikube addons enable ingress
# metrics server gathers resource usage and is nice to have
minikube addons enable metrics-server
# the Kubernetes dashboard
minikube addons enable dashboard
# registry can be used to push images (we'll see what this means later)
minikube addons enable registry

A huger minikube start command may look as follows. Note that CPU and MEMORY may already be limitted by your Docker runtime:

minikube start --driver='docker' --addons='dashboard,ingress,metrics-server,registry' \
--memory='7g' --cpus='4'
# for quick ramp up
helm upgrade --install --create-namespace cert-manager jetstack/cert-manager -n cert-manager -f ./helm_cert-manager.yaml
kubectl apply -f cert-manager_issuer.yaml

Client tools

When a cluster is running we could already send commands to the cluster. To do so, various client tools are required. At least kubectl and helm. K9s and kubens make it easier and helm with switching between clusters.

  • kubectl (equivalent to the docker cli)
    • Visit this page to install kubectl locally.
    • Alternatively you can use minikube kubectl and use a alias for this on the CLI.
  • helm (install "packages" on Kubernetes)
    • Visit this page to install helm locally.
    • For those who have trust: curl https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash
  • k9s (ncurse client application to quickly operate your cluster)
    • Download the binary and execute, for example from github.com.
    • Detailed instructions here.
  • kubectx & kubens
    • Installation is explained here.

(optional) docker cli

Building containers is not required as part of this workshop. But you may want to try it out. Building containers can be done for example with the Docker Engine. Note: There is also Docker Desktop available which is free to use for private use, probably has more comfort, but is also a heavyweight application. Docker Engine is recommended.

To install Docker Engine follow these instructions: https://docs.docker.com/engine/

(optional) reset setup and advanced functions

It might happen that you get stuck during trying out things. This might be caused by your setup simply being completely messed up. But not worries. A fresh cluster can be set up in a minute.

# delete the cluster and recreate a fresh one
minikube delete && minikube start

But might as well get bored by the simple setup and are striving for more. In that case minikube offers a lot of extended features.

  • You may want to take a look at the minikube dashboard, simply execute minikube dashboard.
  • Change the default settings how much cpu and memory your cluster is assigned, see docs.
  • You may want to install one of the offered addons. List with minukube addons list and enable with minikube addon enable youraddonname. Some addons might need configuration.
  • You may want to try out a tutorial as offered on the minikube website.
  • You may want to start your cluster with an alternative driver.

Verify installation

The following commands may be used to verify that everything is working as expected.

# verify minikube
minikube status
# Verify tools are installed with help commands
kuebctl -h
helm -h
k9s -h
# kubectl - query for all running containers in cluster
# (-A means: In all namespaces.)
kubectl get pods -A
# query all installed helm charts in cluster (no output expected)
helm list -A
# start k9s
k9s
# ":po" && "0" => check all pods
# ctrl +c => to close