Skip to content

Github Actions Lab - CI

The goal of this exercise is to show you how to create a CI/CD Pipeline for containerized applications.
For this Lab, we will use the automation platform of Github, called Github Actions.

At the end of the exercise, you should have added a Github Actions YAML file to your repository and on each commit and push to the master branch, a new Docker image should be built automatically, and pushed to dockerhub.

Danger

Performing this lab is absolutely necessary for most of the following exercises.

CI / CD ?

We always refer to the automatization of code source build and deployment with the terms CI/CD, for Continuous Integration and Continuous Deployment. Several tools are able to help us automatize all the steps from the source code to the project running in production.

Warning

  • Continuous Integration is automating the Build steps and the Testing part, to obtain a validated delivrable package.
  • CD can mean : Continuous Delivery or Continuous Deployment. The main difference is that Delivery means : we do not push to prod automatically. Here is a good sum up made by AWS CI/CD teams

CICD

Basic Action

First, let's create a basic action, doing a simple Echo from Mona, the Octocat.

  • Create the .github/workflows directory needed to use github actions.
  • Create the file build.yaml in here.

Compile, Test & Package the Code

Since we already did the job in maven, the easiest is to reuse this in an ubuntu based action with Java enabled, and some mvn steps.

  • Update the build.yaml file : make it clean, give proper names to steps, remove the basic step ...

    Tip

    mvn is embededd and configured in ubuntu:latest github-action container.

    Yet, java need to be configured :

    ...
  - name: Set up JDK 1.8
    uses: actions/setup-java@v1
    with:
      java-version: 1.8
...

    Warning

    Don't forget to checkout the code of your µS inside the github action container at the begining 😉

    Also, choose the most relevant trigger : On push ? On pull request ? On which branch ?

Build & Push the Image

Let's simply use an existing action provided by DockerHub.

  • Login to dockerhub and create a repository for your image : <your_docker_id>/cloudcomputing
  • Update the file build.yaml again.
  • Use the Github Action provided by Docker (github link)

  • Push your image with a unique tag to hub.docker.com (use the GITHUB_SHA environment variable)

    Warning

    Don't forget to create your DockerHub credentials as Github Secrets in your repo settings !!!

  • Check your image has been pushed to your DockerHub Account Registry.

    Success

    Congratulation, your Docker image is ready to be used by anybody !

Followup Exercise

Tip

You can setup a build notification on #Slack ! voxmedia/github-action-slack-notify-build .

Scopes used are :

chat:write

channels:read

groups:read

Tip

You can setup an online Documentation using Github Pages !

Deploy MkDocs.