With this action, you can run your pipelines on GitLab for your GitHub project.

This action takes care of the following things:

• Pushing the relevant code to GitLab
• Starting a pipeline on the relevant code
• Hosting a GitLab Runner
• Waiting for the GitLab pipeline to finish
• Removing the code from GitLab after the pipeline finished
• Failing the workflow if the GitLab pipeline did not complete successful.

1. Create a project on GitLab [Create project]
2. Create a personal access token and save it in the secrets on GitHub (A group or project access token also works) [Create personal access token] [Create a secret]
3. Create a runner and save the runner token in the secrets on GitHub (this step can be skipped if you want to use shared runners) [Create runner] [Create a secret]

Create a workflow with the following content.

Replace <project-id> with the ID of the project you created in step one. The snippet assumes that the secret from step 2 and 3 were created with the name GL_API_TOKEN and GL_RUNNER_TOKEN.

If your project is hosted on gitlab.com, you can omit the GL_SERVER_URL configuration.

name: GitLab

on:
  push:
  pull_request:

jobs:
  pipeline:
    runs-on: ubuntu-latest
    steps:
      - uses: Taucher2003/[email protected]
        name: Run pipeline
        with:
          GL_SERVER_URL: https://gitlab.com
          GL_PROJECT_ID: '<project-id>'
          GL_RUNNER_TOKEN: ${{ secrets.GL_RUNNER_TOKEN }}
          GL_API_TOKEN: ${{ secrets.GL_API_TOKEN }}

You can define variables that should be passed to the GitLab pipeline.

Every environment variable starting with GLPA_ will be passed to the GitLab pipeline with the GLPA_ prefix removed.

name: GitLab

on:
  push:
  pull_request:

jobs:
  pipeline:
    runs-on: ubuntu-latest
    steps:
      - uses: Taucher2003/[email protected]
        name: Run pipeline
        with:
          GL_SERVER_URL: https://gitlab.com
          GL_PROJECT_ID: '<project-id>'
          GL_RUNNER_TOKEN: ${{ secrets.GL_RUNNER_TOKEN }}
          GL_API_TOKEN: ${{ secrets.GL_API_TOKEN }}
        env:
          GLPA_GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

By adding the SHOW_JOB_LOGS to the input section (with), you can show the job logs from the GitLab pipeline in the output of the GitHub Action Run.

Available options are none, failures and all.
When using none or not specifying the option, no job logs will be shown.
With failures, the job logs of failed jobs will be shown.
all shows the job log of all jobs in the pipeline.

GitHub does not pass secrets to actions triggered by pull requests from forks. This causes the GL_API_TOKEN and GL_RUNNER_TOKEN to be empty when triggered from a fork.

You can work around that by using the pull_request_target trigger. Because this triggers the workflow against the base branch instead of the head branch, you need to set overrides for the action to checkout the code from the PR instead of the base branch.

Before adopting this solution, think about the security implications. You can read more about them in the GitHub documentation.

If the GitLab project contains CI/CD variables, they can be extracted from an external contributor by just opening a PR on GitHub.

You can work around that with an environment that requires a review to deploy. That will hold the job until it is approved by a repository collaborator.

name: GitLab

on:
   pull_request_target:

jobs:
   pipeline:
      runs-on: ubuntu-latest
      steps:
         - uses: Taucher2003/[email protected]
           name: Run pipeline
           with:
              GL_SERVER_URL: https://gitlab.com
              GL_PROJECT_ID: '<project-id>'
              GL_RUNNER_TOKEN: ${{ secrets.GL_RUNNER_TOKEN }}
              GL_API_TOKEN: ${{ secrets.GL_API_TOKEN }}
              OVERRIDE_GITHUB_SHA: ${{ github.event.pull_request.head.sha }}
              OVERRIDE_GITHUB_REF_NAME: ${{ github.event.pull_request.head.ref }}