Multi-project pipelines

Introduced in GitLab Premium 9.3.

When you set up GitLab CI/CD across multiple projects, you can visualize the entire pipeline, including all cross-project inter-dependencies.

Overview

GitLab CI/CD is a powerful continuous integration tool that works not only per project, but also across projects. When you configure GitLab CI for your project, you can visualize the stages of your jobs on a pipeline graph.

Multi-project pipeline graph

In the Merge Request Widget, multi-project pipeline mini-graphs are displayed, and when hovering or tapping (on touchscreen devices) they will expand and be shown adjacent to each other.

Multi-project mini graph

Multi-project pipelines are useful for larger products that require cross-project inter-dependencies, such as those adopting a microservices architecture.

For a demonstration of how cross-functional development teams can use cross-pipeline triggering to trigger multiple pipelines for different microservices projects, see Cross-project Pipeline Triggering and Visualization.

Use cases

Let’s assume you deploy your web app from different projects in GitLab:

  • One for the free version, which has its own pipeline that builds and tests your app
  • One for the paid version add-ons, which also pass through builds and tests
  • One for the documentation, which also builds, tests, and deploys with an SSG

With Multi-Project Pipelines, you can visualize the entire pipeline, including all stages of builds and tests for the three projects.

Triggering multi-project pipelines through API

When you use the CI_JOB_TOKEN to trigger pipelines, GitLab recognizes the source of the job token, and thus internally ties these pipelines together, allowing you to visualize their relationships on pipeline graphs.

These relationships are displayed in the pipeline graph by showing inbound and outbound connections for upstream and downstream pipeline dependencies.

Creating multi-project pipelines from .gitlab-ci.yml

Triggering a downstream pipeline using a bridge job

Before GitLab 11.8, it was necessary to implement a pipeline job that was responsible for making the API request to trigger a pipeline in a different project.

In GitLab 11.8, GitLab provides a new CI/CD configuration syntax to make this task easier, and avoid needing GitLab Runner for triggering cross-project pipelines. The following illustrates configuring a bridge job:

rspec:
  stage: test
  script: bundle exec rspec

staging:
  variables:
    ENVIRONMENT: staging
  stage: deploy
  trigger: my/deployment

In the example above, as soon as rspec job succeeds in the test stage, the staging bridge job is going to be started. The initial status of this job will be pending. GitLab will create a downstream pipeline in the my/deployment project and, as soon as the pipeline gets created, the staging job will succeed. my/deployment is a full path to that project.

The user that created the upstream pipeline needs to have access rights to the downstream project (my/deployment in this case). If a downstream project can not be found, or a user does not have access rights to create pipeline there, the staging job is going to be marked as failed.

Caution: In the example, staging will be marked as succeeded as soon as a downstream pipeline gets created. If you want to display the downstream pipeline’s status instead, see Mirroring status from triggered pipeline.
Note: Bridge jobs do not support every configuration entry that a user can use in the case of regular jobs. Bridge jobs will not to be picked by a Runner, thus there is no point in adding support for script, for example. If a user tries to use unsupported configuration syntax, YAML validation will fail upon pipeline creation.

Specifying a downstream pipeline branch

It is possible to specify a branch name that a downstream pipeline will use:

rspec:
  stage: test
  script: bundle exec rspec

staging:
  stage: deploy
  trigger:
    project: my/deployment
    branch: stable-11-2

Use:

  • The project keyword to specify the full path to a downstream project.
  • The branch keyword to specify the name of a branch in the project specified by project. From GitLab 12.4, variable expansion is supported.

GitLab will use a commit that is currently on the HEAD of the branch when creating a downstream pipeline.

Passing variables to a downstream pipeline

Sometimes you might want to pass variables to a downstream pipeline. You can do that using the variables keyword, just like you would when defining a regular job.

rspec:
  stage: test
  script: bundle exec rspec

staging:
  variables:
    ENVIRONMENT: staging
  stage: deploy
  trigger: my/deployment

The ENVIRONMENT variable will be passed to every job defined in a downstream pipeline. It will be available as an environment variable when GitLab Runner picks a job.

In the following configuration, the MY_VARIABLE variable will be passed to the downstream pipeline that is created when the trigger-downstream job is queued. This is because trigger-downstream job inherits variables declared in global variables blocks, and then we pass these variables to a downstream pipeline.

variables:
  MY_VARIABLE: my-value

trigger-downstream:
  variables:
    ENVIRONMENT: something
  trigger: my/project

You might want to pass some information about the upstream pipeline using, for example, predefined variables. In order to do that, you can use interpolation to pass any variable. For example:

downstream-job:
  variables:
    UPSTREAM_BRANCH: $CI_COMMIT_REF_NAME
  trigger: my/project

In this scenario, the UPSTREAM_BRANCH variable with a value related to the upstream pipeline will be passed to the downstream-job job, and will be available within the context of all downstream builds.

Tip: Upstream pipelines take precedence over downstream ones. If there are two variables with the same name defined in both upstream and downstream projects, the ones defined in the upstream project will take precedence.

Mirroring status from triggered pipeline

You can mirror the pipeline status from the triggered pipeline to the source bridge job by using strategy: depend. For example:

trigger_job:
  trigger:
    project: my/project
    strategy: depend

Mirroring status from upstream pipeline

You can mirror the pipeline status from an upstream pipeline to a bridge job by using the needs:pipeline keyword. The latest pipeline status from master is replicated to the bridge job.

Example:

upstream_bridge:
  stage: test
  needs:
    pipeline: other/project

Limitations

Because bridge jobs are a little different to regular jobs, it is not possible to use exactly the same configuration syntax here, as one would normally do when defining a regular job that will be picked by a runner.

Some features are not implemented yet. For example, support for environments.

Configuration keywords available for bridge jobs are:

  • trigger (to define a downstream pipeline trigger)
  • stage
  • allow_failure
  • only and except
  • when (only with on_success, on_failure, and always values)
  • extends