This tutorial gives you hands on docker, kubernetes, helm and skaffold experience. This is done using a very simple application: vue on the frontend and fastapi python on the backend. The only functionality is to get the time!
You can set up a hetzner cloud environment if you don't want to work on a local machine. You can do this using the pulumi stack here.
As an alternative, you can work locally. Make sure you have a recent linux distribution with docker installed (eg ubuntu 20.04 with the docker.io package). In order to follow this tutorial in a comfortable way, adhere to:
- At least 8GB of ram (16 will be even more comfortable, especially if you want to run a heavy IDE)
- At least 40GB of disk space. When disk space is getting low, kubernetes taints your nodes and your pods won't run anymore!
- Don't use a hard disk, use a SSD or kubernetes will be horribly slow. Preferably a nvme ssd, not a SATA one.
If you decided to work remotely, you probably want to use Visual studio code remote SSH extension
To get started with the tutorial, clone this repository locally:
git clone https://github.com/Kapernikov/skaffold-helm-tutorial
Then, work in this directory skaffold-helm-tutorial
for all the rest of the tutorial.
This tutorial assumes some knowledge beforehand. You will have trouble following if you don't have the following:
- A basic understanding of docker. (You must have built a docker image and have written a Dockerfile already). There are lots of tutorials on docker.
- Some knowledge on linux shell (bash) scripting. Linux shell is used everywhere to glue all kind of stuff together. Go here for a tutorial on using the linux terminal, and here for a tutorial on writing shell scripts (however, you won't need all of the advanced stuff mentioned in that tutorial, so a couple of chapters will do fine).
This tutorial assumes the use of linux. You can also use a virtual machine on windows or WSL2 (avoid WSL1, it won't work). For WSL2, you need to take into account some caveats. See specific instructions here.
This tutorial requires docker! So you need to install docker on your linux machine. On ubuntu, it's as simple as
sudo apt install docker.io
This will install the bundled version of docker, which is not totally up to date. If you want to use newer features (like buildx) you need to install docker ce according to the official instructions.
In addition to this, some frequently used tools are also needed:
sudo apt install wget curl vim zip git
-
Optional: get set up to do this tutorial on a remote machine
-
Chapter 1: Installing k3s
-
Chapter 2: Dockerize our API
-
Chapter 3: Dockerize our frontend app
-
Chapter 4: Kubernetes
-
Chapter 5: Helm
-
Chapter 6: Skaffold
-
Chapter 8: Ingress
-
Chapter 9: Working with 3rd party software
-
Chapter 10: gitops