Essential tools #

In order for us to get started and be efficient from the get go we need to prepare our environment with the tools we need in order to learn and explore Kubernetes (will from here on be abbreviated as k8s).

Text Editor #

You will need a text editor, use the one you know the best. If you don’t have any in particular I would recommend using Visual Studio Code.

Docker #

The most important tool we will be using is Docker. Docker is used for building and running container images. We will also be using docker for running k8s locally.

For most people Docker Desktop is the best and easiest way of running docker. On some corporate (windows) laptops it might be easier to run docker directly in WSL (I won’t go into detail about how to set this up as instructions might differ from place to place)

If you can I would recommend running docker desktop as this abstracts the complexity of configuring docker with correct settings, networking etc. You can verify that docker has been installed and is running correctly using this command: docker ps this should return the following output:

CONTAINER ID   IMAGE     COMMAND   CREATED   STATUS    PORTS     NAMES

This might look a bit strange but this is a list of the containers you are currently running. In the example above, the list is empty as we haven’t started any containers yet.

If you get an error like

Cannot connect to the Docker daemon at unix:///Users/blacksails/.docker/run/docker.sock. Is the docker daemon running?

This means that docker is not running and you will need to start it. For Docker Desktop this is simply a matter of opening the application.

Kubectl #

In order to interact with kubernetes clusters the main tool is called kubectl. This is a CLI which we use to query state and make changes to our cluster. The kubectl installation instructions can be found here.

In order to test wether you got kubectl installed, you can run the following command: kubectl version. It should show an output similar to the following:

Client Version: v1.33.1
Kustomize Version: v5.6.0
The connection to the server localhost:8080 was refused - did you specify the right host or port?

Don’t mind the message about connection to the server. This is simply because we haven’t set it up to connect to a kubernetes cluster yet (more on this later).

Helm #

Helm is a “package manager” for k8s which aims to make it easier to install things in your k8s cluster. We will be using it through out the workshop and also look into how helm works during the workshop.

You can find the installation instructions here.

Flux #

Flux is an implementation of the GitOps pattern (more on this later). Installation instructions

Go toolchain #

Go is a programming language widely used in the k8s ecosystem. docker and kubectl are actually written in go. But don’t worry you will not need to learn the language in order to complete this course, we will mainly use it in order to easily install tools that we need.

The go installation instructions can be found here.

In addition to the official installation instructions I would recommend that you explicitly set the GOPATH and GOBIN environment variables and add GOBIN to the PATH environment variable.

For most people this means adding a few lines to ~/.zshrc. Open the file and paste the following lines at the bottom:

export GOPATH="$HOME/go"
export GOBIN="$GOPATH/bin"
export PATH="$PATH:$GOBIN"

This tells the go tool chain where to find source code and where to install binaries. By adding GOBIN to the PATH environment variable we ensure that our shell is able to find the tools we install using the go toolchain.

When installing tools using the go tool chain, we will be using the go install <tool> command This will download the source code, compile a binary and place it in the GOBIN directory.

KIND #

In order to easily run kubernetes locally we will be using a tool called kind. The name of this tool is an abbreviation of Kubernetes in Docker (more on this later).

With the go toolchain installed we can easily install kind by running the following command:

go install sigs.k8s.io/[email protected]

Cluster API #

Cluster API is an API for managing kubernetes clusters as kubernetes resources (more on this later).

go install sigs.k8s.io/cluster-api/cmd/clusterctl@latest

Additional tools #

The following are additional optional tools which are easy to install using the go tool chain.

yq #

yq is a small tool which can be used to transform yaml

go install github.com/mikefarah/yq/v4@latest

stern #

stern is a tool which makes it easier to view logs in k8s clusters.

go install github.com/stern/stern@latest

k9s #

k9s is a terminal UI (TUI) for viewing and and manipulating kubernetes clusters. I don’t personally use this, but I know a lot of people like it a lot. Think of it as a dynamic user interface compared to the CLI nature that kubectl has.A

go install github.com/derailed/k9s@latest

kubeseal (Sealed Secrets) #

kubeseal is a CLI for which can be used to encrypt secrets, so that they are safe to store in git. Wether we will have time to dig into this is a bit uncertain.

go install github.com/bitnami-labs/sealed-secrets/cmd/kubeseal@main