Pipelines for the GitLab project

Pipelines for gitlab-org/gitlab (as well as the dev instance's) is configured in the usual .gitlab-ci.yml which itself includes files under .gitlab/ci/ for easier maintenance.

We're striving to dogfood GitLab CI/CD features and best-practices as much as possible.

Minimal test jobs before a merge request is approved

To reduce the pipeline cost and shorten the job duration, before a merge request is approved, the pipeline will run a minimal set of RSpec & Jest tests that are related to the merge request changes.

After a merge request has been approved, the pipeline would contain the full RSpec & Jest tests. This will ensure that all tests have been run before a merge request is merged.

Overview of the GitLab project test dependency

To understand how the minimal test jobs are executed, we need to understand the dependency between GitLab code (frontend and backend) and the respective tests (Jest and RSpec). This dependency can be visualized in the following diagram:

flowchart LR
    subgraph frontend
    fe["Frontend code"]--tested with-->jest
    end
    subgraph backend
    be["Backend code"]--tested with-->rspec
    end
    
    be--generates-->fixtures["frontend fixtures"]
    fixtures--used in-->jest

In summary:

• RSpec tests are dependent on the backend code.
• Jest tests are dependent on both frontend and backend code, the latter through the frontend fixtures.

RSpec minimal jobs

Determining related RSpec test files in a merge request

To identify the minimal set of tests needed, we use the test_file_finder gem, with two strategies:

• dynamic mapping from test coverage tracing (generated via the Crystalball gem) (see where it's used)
• static mapping maintained in the tests.yml file for special cases that cannot be mapped via coverage tracing (see where it's used)

The test mappings contain a map of each source files to a list of test files which is dependent of the source file.

In the detect-tests job, we use this mapping to identify the minimal tests needed for the current merge request.

Exceptional cases

In addition, there are a few circumstances where we would always run the full RSpec tests:

• when the pipeline:run-all-rspec label is set on the merge request
• when the merge request is created by an automation (e.g. Gitaly update or MR targeting a stable branch)
• when the merge request is created in a security mirror
• when any CI config file is changed (i.e. .gitlab-ci.yml or .gitlab/ci/**/*)

Jest minimal jobs

Determining related Jest test files in a merge request

To identify the minimal set of tests needed, we pass a list of all the changed files into jest using the --findRelatedTests option. In this mode, jest would resolve all the dependencies of related to the changed files, which include test files that have these files in the dependency chain.

Exceptional cases

In addition, there are a few circumstances where we would always run the full Jest tests:

• when the pipeline:run-all-jest label is set on the merge request
• when the merge request is created by an automation (e.g. Gitaly update or MR targeting a stable branch)
• when the merge request is created in a security mirror
• when any CI config file is changed (i.e. .gitlab-ci.yml or .gitlab/ci/**/*)
• when any frontend "core" file is changed (i.e. package.json, yarn.lock, babel.config.js, jest.config.*.js, config/helpers/**/*.js)
• when any vendored JavaScript file is changed (i.e. vendor/assets/javascripts/**/*)
• when any backend file is changed (see the patterns list for details)

Fail-fast job in merge request pipelines

To provide faster feedback when a merge request breaks existing tests, we are experimenting with a fail-fast mechanism.

An rspec fail-fast job is added in parallel to all other rspec jobs in a merge request pipeline. This job runs the tests that are directly related to the changes in the merge request.

If any of these tests fail, the rspec fail-fast job fails, triggering a fail-pipeline-early job to run. The fail-pipeline-early job:

• Cancels the currently running pipeline and all in-progress jobs.
• Sets pipeline to have status failed.

For example:

graph LR
    subgraph "prepare stage";
        A["detect-tests"]
    end

    subgraph "test stage";
        B["jest"];
        C["rspec migration"];
        D["rspec unit"];
        E["rspec integration"];
        F["rspec system"];
        G["rspec fail-fast"];
    end

    subgraph "post-test stage";
        Z["fail-pipeline-early"];
    end

    A --"artifact: list of test files"--> G
    G --"on failure"--> Z

The rspec fail-fast is a no-op if there are more than 10 test files related to the merge request. This prevents rspec fail-fast duration from exceeding the average rspec job duration and defeating its purpose.

This number can be overridden by setting a CI/CD variable named RSPEC_FAIL_FAST_TEST_FILE_COUNT_THRESHOLD.

Test jobs

We have dedicated jobs for each testing level and each job runs depending on the changes made in your merge request. If you want to force all the RSpec jobs to run regardless of your changes, you can add the pipeline:run-all-rspec label to the merge request.

WARNING: Forcing all jobs on docs only related MRs would not have the prerequisite jobs and would lead to errors

Test suite parallelization

Our current RSpec tests parallelization setup is as follows:

1. The retrieve-tests-metadata job in the prepare stage ensures we have a knapsack/report-master.json file:
   • The knapsack/report-master.json file is fetched from the latest main pipeline which runs update-tests-metadata (for now it's the 2-hourly scheduled master pipeline), if it's not here we initialize the file with {}.
2. Each [rspec|rspec-ee] [unit|integration|system|geo] n m job are run with knapsack rspec and should have an evenly distributed share of tests:
   • It works because the jobs have access to the knapsack/report-master.json since the "artifacts from all previous stages are passed by default".
   • the jobs set their own report path to "knapsack/${TEST_TOOL}_${TEST_LEVEL}_${DATABASE}_${CI_NODE_INDEX}_${CI_NODE_TOTAL}_report.json".
   • if knapsack is doing its job, test files that are run should be listed under Report specs, not under Leftover specs.
3. The update-tests-metadata job (which only runs on scheduled pipelines for the canonical project takes all the knapsack/rspec*_pg_*.json files and merge them all together into a single knapsack/report-master.json file that is saved as artifact.

After that, the next pipeline uses the up-to-date knapsack/report-master.json file.

Flaky tests

Tests that are known to be flaky are skipped unless the $SKIP_FLAKY_TESTS_AUTOMATICALLY variable is set to false or if the ~"pipeline:run-flaky-tests" label is set on the MR.

Monitoring

The GitLab test suite is monitored for the main branch, and any branch that includes rspec-profile in their name.

Logging

• Rails logging to log/test.log is disabled by default in CI for performance reasons. To override this setting, provide the RAILS_ENABLE_TEST_LOG environment variable.

Review app jobs

Consult the Review Apps dedicated page for more information.

If you want to force a Review App to be deployed regardless of your changes, you can add the pipeline:run-review-app label to the merge request.

As-if-FOSS jobs

The * as-if-foss jobs run the GitLab test suite "as if FOSS", meaning as if the jobs would run in the context of the gitlab-org/gitlab-foss project. These jobs are only created in the following cases:

• when the pipeline:run-as-if-foss label is set on the merge request
• when the merge request is created in the gitlab-org/security/gitlab project
• when any CI config file is changed (i.e. .gitlab-ci.yml or .gitlab/ci/**/*)

The * as-if-foss jobs are run in addition to the regular EE-context jobs. They have the FOSS_ONLY='1' variable set and get the ee/ folder removed before the tests start running.

The intent is to ensure that a change doesn't introduce a failure after the gitlab-org/gitlab project is synced to the gitlab-org/gitlab-foss project.

As-if-JH jobs

The * as-if-jh jobs run the GitLab test suite "as if JiHu", meaning as if the jobs would run in the context of the gitlab-jh/gitlab project. These jobs are only created in the following cases:

• when the pipeline:run-as-if-jh label is set on the merge request
• when the pipeline:run-all-rspec label is set on the merge request
• when any code or backstage file is changed
• when any startup CSS file is changed

The * as-if-jh jobs are run in addition to the regular EE-context jobs. The jh/ folder is added before the tests start running.

The intent is to ensure that a change doesn't introduce a failure after the gitlab-org/gitlab project is synced to the gitlab-jh/gitlab project.

Corresponding JH branch

You can create a corresponding JH branch on the gitlab-jh/gitlab project by appending -jh to the branch name. If a corresponding JH branch is found, * as-if-jh jobs grab the jh folder from the respective branch, rather than from the default branch.

undercover RSpec test

Introduced in GitLab 14.6.

The rspec:undercoverage job runs undercover to detect, and fail if any changes introduced in the merge request has zero coverage.

The rsepc:undercoverage job obtains coverage data from the rspec:coverage job.

In the event of an emergency, or false positive from this job, add the pipeline:skip-undercoverage label to the merge request to allow this job to fail.

PostgreSQL versions testing

Our test suite runs against PG12 as GitLab.com runs on PG12 and Omnibus defaults to PG12 for new installs and upgrades.

We do run our test suite against PG11 and PG13 on nightly scheduled pipelines.

We also run our test suite against PG11 upon specific database library changes in MRs and main pipelines (with the rspec db-library-code pg11 job).

Current versions testing

Long-term plan

We follow the PostgreSQL versions shipped with Omnibus GitLab:

Pipelines types for merge requests

In general, pipelines for an MR fall into one of the following types (from shorter to longer), depending on the changes made in the MR:

• Documentation pipeline: For MRs that touch documentation.
• Backend pipeline: For MRs that touch backend code.
• Frontend pipeline: For MRs that touch frontend code.
• End-to-end pipeline: For MRs that touch code in the qa/ folder.

A "pipeline type" is an abstract term that mostly describes the "critical path" (i.e. the chain of jobs for which the sum of individual duration equals the pipeline's duration). We use these "pipeline types" in metrics dashboards in order to detect what types and jobs need to be optimized first.

An MR that touches multiple areas would be associated with the longest type applicable. For instance, an MR that touches backend and frontend would fall into the "Frontend" pipeline type since this type takes longer to finish than the "Backend" pipeline type.

We use the rules: and needs: keywords extensively to determine the jobs that need to be run in a pipeline. Note that an MR that includes multiple types of changes would have a pipelines that include jobs from multiple types (for example, a combination of docs-only and code-only pipelines).

Following are graphs of the critical paths for each pipeline type. Jobs that aren't part of the critical path are ommitted.

Documentation pipeline

Reference pipeline.

graph LR
  classDef criticalPath fill:#f66;

  1-3["docs-lint links (5 minutes)"];
  class 1-3 criticalPath;
  click 1-3 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=8356757&udv=0"

Backend pipeline

Reference pipeline.

graph RL;
  classDef criticalPath fill:#f66;

  1-3["compile-test-assets (6 minutes)"];
  class 1-3 criticalPath;
  click 1-3 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914317&udv=0"
  1-6["setup-test-env (4 minutes)"];
  click 1-6 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914315&udv=0"
  1-14["retrieve-tests-metadata"];
  click 1-14 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=8356697&udv=0"
  1-15["detect-tests"];
  click 1-15 "https://app.periscopedata.com/app/gitlab/652085/EP---Jobs-Durations?widget=10113603&udv=1005715"

  2_5-1["rspec & db jobs (24 minutes)"];
  class 2_5-1 criticalPath;
  click 2_5-1 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations"
  2_5-1 --> 1-3 & 1-6 & 1-14 & 1-15;

  3_2-1["rspec:coverage (5.35 minutes)"];
  class 3_2-1 criticalPath;
  click 3_2-1 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=7248745&udv=0"
  3_2-1 -.->|"(don't use needs<br/>because of limitations)"| 2_5-1;

  4_3-1["rspec:undercoverage (3.5 minutes)"];
  class 4_3-1 criticalPath;
  click 4_3-1 "https://app.periscopedata.com/app/gitlab/652085/EP---Jobs-Durations?widget=13446492&udv=1005715"
  4_3-1 --> 3_2-1;

Frontend pipeline

Reference pipeline.

graph RL;
  classDef criticalPath fill:#f66;

  1-2["build-qa-image (2 minutes)"];
  click 1-2 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914325&udv=0"
  1-5["compile-production-assets (16 minutes)"];
  class 1-5 criticalPath;
  click 1-5 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914312&udv=0"

  2_3-1["build-assets-image (1.3 minutes)"];
  class 2_3-1 criticalPath;
  2_3-1 --> 1-5

  2_6-1["start-review-app-pipeline (49 minutes)"];
  class 2_6-1 criticalPath;
  click 2_6-1 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations"
  2_6-1 --> 2_3-1 & 1-2;

End-to-end pipeline

Reference pipeline.

graph RL;
  classDef criticalPath fill:#f66;

  1-2["build-qa-image (2 minutes)"];
  click 1-2 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914325&udv=0"
  1-5["compile-production-assets (16 minutes)"];
  class 1-5 criticalPath;
  click 1-5 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914312&udv=0"
  1-15["detect-tests"];
  click 1-15 "https://app.periscopedata.com/app/gitlab/652085/EP---Jobs-Durations?widget=10113603&udv=1005715"

  2_3-1["build-assets-image (1.3 minutes)"];
  class 2_3-1 criticalPath;
  2_3-1 --> 1-5

  2_4-1["package-and-qa (102 minutes)"];
  class 2_4-1 criticalPath;
  click 2_4-1 "https://app.periscopedata.com/app/gitlab/652085/Engineering-Productivity---Pipeline-Build-Durations?widget=6914305&udv=0"
  2_4-1 --> 1-2 & 2_3-1 & 1-15;

CI configuration internals

Workflow rules

Pipelines for the GitLab project are created using the workflow:rules keyword feature of the GitLab CI/CD.

Pipelines are always created for the following scenarios:

• main branch, including on schedules, pushes, merges, and so on.
• Merge requests.
• Tags.
• Stable, auto-deploy, and security branches.

Pipeline creation is also affected by the following CI/CD variables:

• If $FORCE_GITLAB_CI is set, pipelines are created.
• If $GITLAB_INTERNAL is not set, pipelines are not created.

No pipeline is created in any other cases (for example, when pushing a branch with no MR for it).

The source of truth for these workflow rules is defined in .gitlab-ci.yml.

Default image

The default image is defined in .gitlab-ci.yml.

It includes Ruby, Go, Git, Git LFS, Chrome, Node, Yarn, PostgreSQL, and Graphics Magick.

The images used in our pipelines are configured in the gitlab-org/gitlab-build-images project, which is push-mirrored to gitlab/gitlab-build-images for redundancy.

The current version of the build images can be found in the "Used by GitLab section".

Default variables

In addition to the predefined CI/CD variables, each pipeline includes default variables defined in .gitlab-ci.yml.

Stages

The current stages are:

• sync: This stage is used to synchronize changes from gitlab-org/gitlab to gitlab-org/gitlab-foss.
• prepare: This stage includes jobs that prepare artifacts that are needed by jobs in subsequent stages.
• build-images: This stage includes jobs that prepare Docker images that are needed by jobs in subsequent stages or downstream pipelines.
• fixtures: This stage includes jobs that prepare fixtures needed by frontend tests.
• lint: This stage includes linting and static analysis jobs.
• test: This stage includes most of the tests, and DB/migration jobs.
• post-test: This stage includes jobs that build reports or gather data from the test stage's jobs (for example, coverage, Knapsack metadata, and so on).
• review-prepare: This stage includes a job that build the CNG images that are later used by the (Helm) Review App deployment (see Review Apps for details).
• review: This stage includes jobs that deploy the GitLab and Docs Review Apps.
• dast: This stage includes jobs that run a DAST full scan against the Review App that is deployed in stage review.
• qa: This stage includes jobs that perform QA tasks against the Review App that is deployed in stage review.
• post-qa: This stage includes jobs that build reports or gather data from the qa stage's jobs (for example, Review App performance report).
• pages: This stage includes a job that deploys the various reports as GitLab Pages (for example, coverage-ruby, and webpack-report (found at https://gitlab-org.gitlab.io/gitlab/webpack-report/, but there is an issue with the deployment).
• notify: This stage includes jobs that notify various failures to Slack.

Dependency Proxy

Some of the jobs are using images from Docker Hub, where we also use ${GITLAB_DEPENDENCY_PROXY} as a prefix to the image path, so that we pull images from our Dependency Proxy.

${GITLAB_DEPENDENCY_PROXY} is a group CI/CD variable defined in gitlab-org as ${CI_DEPENDENCY_PROXY_GROUP_IMAGE_PREFIX}/. This means when we use an image defined as:

image: ${GITLAB_DEPENDENCY_PROXY}alpine:edge

Projects in the gitlab-org group pull from the Dependency Proxy, while forks that reside on any other personal namespaces or groups fall back to Docker Hub unless ${GITLAB_DEPENDENCY_PROXY} is also defined there.

Common job definitions

Most of the jobs extend from a few CI definitions defined in .gitlab/ci/global.gitlab-ci.yml that are scoped to a single configuration keyword.

rules, if: conditions and changes: patterns

We're using the rules keyword extensively.

All rules definitions are defined in rules.gitlab-ci.yml, then included in individual jobs via extends.

The rules definitions are composed of if: conditions and changes: patterns, which are also defined in rules.gitlab-ci.yml and included in rules definitions via YAML anchors

if: conditions

changes: patterns

Performance

Interruptible pipelines

By default, all jobs are interruptible, except the dont-interrupt-me job which runs automatically on main, and is manual otherwise.

If you want a running pipeline to finish even if you push new commits to a merge request, be sure to start the dont-interrupt-me job before pushing.

Caching strategy

1. All jobs must only pull caches by default.
2. All jobs must be able to pass with an empty cache. In other words, caches are only there to speed up jobs.
3. We currently have several different cache definitions defined in .gitlab/ci/global.gitlab-ci.yml, with fixed keys:
   • .setup-test-env-cache
   • .ruby-cache
   • .rails-cache
   • .static-analysis-cache
   • .rubocop-cache
   • .coverage-cache
   • .danger-review-cache
   • .qa-cache
   • .yarn-cache
   • .assets-compile-cache (the key includes ${NODE_ENV} so it's actually two different caches).
4. These cache definitions are composed of multiple atomic caches.
5. Only the following jobs, running in 2-hourly scheduled pipelines, are pushing (that is, updating) to the caches:
   • update-setup-test-env-cache, defined in .gitlab/ci/rails.gitlab-ci.yml.
   • update-gitaly-binaries-cache, defined in .gitlab/ci/rails.gitlab-ci.yml.
   • update-rubocop-cache, defined in .gitlab/ci/rails.gitlab-ci.yml.
   • update-qa-cache, defined in .gitlab/ci/qa.gitlab-ci.yml.
   • update-assets-compile-production-cache, defined in .gitlab/ci/frontend.gitlab-ci.yml.
   • update-assets-compile-test-cache, defined in .gitlab/ci/frontend.gitlab-ci.yml.
   • update-yarn-cache, defined in .gitlab/ci/frontend.gitlab-ci.yml.
   • update-storybook-yarn-cache, defined in .gitlab/ci/frontend.gitlab-ci.yml.
6. These jobs can also be forced to run in merge requests with the pipeline:update-cache label (this can be useful to warm the caches in a MR that updates the cache keys).

Artifacts strategy

We limit the artifacts that are saved and retrieved by jobs to the minimum in order to reduce the upload/download time and costs, as well as the artifacts storage.

Git fetch caching

Because GitLab.com uses the pack-objects cache, concurrent Git fetches of the same pipeline ref are deduplicated on the Gitaly server (always) and served from cache (when available).

This works well for the following reasons:

• The pack-objects cache is enabled on all Gitaly servers on GitLab.com.
• The CI/CD Git strategy setting for gitlab-org/gitlab is Git clone, causing all jobs to fetch the same data, which maximizes the cache hit ratio.
• We use shallow clone to avoid downloading the full Git history for every job.

Pre-clone step

NOTE: We no longer use this optimization for gitlab-org/gitlab because the pack-objects cache allows Gitaly to serve the full CI/CD fetch traffic now. See Git fetch caching.

The pre-clone step works by using the CI_PRE_CLONE_SCRIPT variable defined by GitLab.com shared runners.

Return to Development documentation