Skip to content

automation_exercise.md

Latest commit

 

History

History
37 lines (23 loc) · 3.12 KB

automation_exercise.md

File metadata and controls

37 lines (23 loc) · 3.12 KB
Raw

Automating Workflows with GitHub Actions

In this exercise, we create automated workflows and pipelines based on GitHub Actions. We are again working with our beloved diffusion Python code.

Deadline: Wednesday, January 28th, 2025, 09:00

Preparations

Import (not fork) the automation exercise repository in your own account/namespace on GitHub.

Note: We cannot work with forks here because GitHub Actions may not work in pull requests without explicit approval of the owner of the target repository. It is also easier to add badges (see below) to the README.md if the repository is under your control.

Task descriptions

The CI workflow should be triggered by push events and should have three jobs:

  1. Name: style_check. Check the code style of the Python files in the repository using black. This check should fail if black finds any file to reformat.
  2. Name: test. Run tests while also collecting coverage information via pytest and the coverage package. The coverage information (stored in the file .coverage) has to be stored and handed over to the next step.
    • The intermediate coverage information (.coverage) should be kept for one day. This is the minimum amount of time an artifact must be stored.
    • Adding a .gitignore file helps to not accidentally commit the temporary coverage files.
  3. Name: coverage_report. Create a coverage report (coverage report) in the terminal and afterwards also convert the coverage information into XML format (coverage xml). For this you have to reuse the coverage information (i.e. .coverage) from the previous step.
    • Running coverage report allows us to inspect the coverage directly in the workflow's output. Saving the XML file allows us to analyze the coverage in more detail with other tools if needed.
    • The resulting file coverage.xml should be kept for 14 days.

Once the workflow runs successfully, add a GitHub workflow status badge for your workflow to the README.md. Label this badge with "SSE CI/CD".

Submission

  • Submit your solution via an issue in the upstream automation exercise repository. The issue should be named [USERNAME] Automation exercise, e.g., [uekermbn] Automation exercise. Please use your GitLab username here. Add a link to your GitHub repository in the issue. If necessary, add further explanations in the issue description.

Bonus: GitLab CI/CD

Realize the same workflows in GitLab CI/CD :)

  • Create a new project under your namespace and push your existing automation repository to this remote.
  • Create a VM on bwCloud and create a GitLab Runner with a Docker executor there. Register the runner in your repository.
  • Add a link to this repository in the submission issue above and, if necessary, explain what you did. Be sure that your repository is visible to us (either public or add gchourdakis).