Text of docs.gitlab.com › ci › migration › github_actions as Styx crawled it on Oct 8, 2026. The page may have changed since. Only its words are kept, without pictures or layout, and only the first part of a long page.

Open the live page

Migrate from GitHub Actions | GitLab Docs

/ Migrate from GitHub Actions Tier : Free, Premium, Ultimate Offering : GitLab.com, GitLab Self-Managed, GitLab Dedicated If you’re migrating from GitHub Actions to GitLab CI/CD, you can create CI/CD pipelines that replicate and enhance your GitHub Action workflows. You can do this by hand, or you can use your agent of choice with the GitHub Actions to GitLab CI/CD agent skill Key similarities and differences GitHub Actions and GitLab CI/CD are both used to generate pipelines to automate building, testing, and deploying your code.

Both share similarities including: CI/CD functionality has direct access to the code stored in the project repository. Pipeline configurations written in YAML and stored in the project repository. Pipelines are configurable and can run in different stages. Jobs can each use a different container image. Additionally, there are some important differences between the two: GitHub has a marketplace for downloading 3rd-party actions, which might require additional support or licenses.

GitLab maintains and supports all features in house, and some 3rd-party integrations are accessible through templates. GitLab provides a built-in container registry. GitLab has native Kubernetes deployment support. GitLab provides granular security policies. Comparison of features and concepts Many GitHub features and concepts have equivalents in GitLab that offer the same functionality. Configuration file GitHub Actions can be configured with a workflow YAML file .

GitLab CI/CD uses a .gitlab-ci.yml YAML file by default. For example, in a GitHub Actions workflow file: yaml on : [ push] jobs : hello : runs-on : ubuntu-latest steps : - run : echo "Hello World" The equivalent GitLab CI/CD .gitlab-ci.yml file would be: yaml stages : - hello hello : stage : hello script : - echo "Hello World" GitHub Actions workflow syntax A GitHub Actions configuration is defined in a workflow YAML file using specific keywords.

GitLab CI/CD has similar functionality, also usually configured with YAML keywords. GitHub GitLab Explanation env variables env defines the variables set in a workflow, job, or step. GitLab uses variables to define CI/CD variables at the global or job level. Variables can also be added in the UI. jobs stages jobs groups together all the jobs that run in the workflow. GitLab uses stages to group jobs together. on Not applicable on defines when a workflow is triggered.

GitLab is integrated tightly with Git, so SCM polling options for triggers are not needed, but can be configured per job if required. run Not applicable The command to execute in the job. GitLab uses a YAML array under the script keyword, one entry for each command to execute. runs-on tags runs-on defines the GitHub runner that a job must run on. GitLab uses tags to select a runner. steps script steps groups together all the steps that run in a job.

GitLab uses script to group together all the commands run in a job. uses include uses defines what GitHub Action to be added to a step . GitLab uses include to add configuration from other files to a job. Common configurations This section goes over commonly used CI/CD configurations, showing how they can be converted from GitHub Actions to GitLab CI/CD. GitHub Action workflows generate automated CI/CD jobs that are triggered when certain events take place, for example pushing a new commit.

A GitHub Action workflow is a YAML file defined in the .github/workflows directory located in the root of the repository. The GitLab equivalent is the .gitlab-ci.yml configuration file, which also resides in the repository’s root directory. Jobs Jobs are a set of commands that run in a set sequence to achieve a particular result, for example building a container or deploying to production. For example, this GitHub Actions workflow builds a container then deploys it to production.

The jobs run sequentially, because the deploy job depends on the build job: yaml on : [ push] jobs : build : runs-on : ubuntu-latest container : golang:alpine steps : - run : apk update - run : go build -o bin/hello - uses : actions/upload-artifact@v3 with : name : hello path : bin/hello retention-days : 7 deploy : if : contains( github.ref, 'staging') runs-on : ubuntu-latest container : golang:alpine steps : - uses : actions/download-artifact@v3 with : name : hello - run : echo "Deploying to Staging" - run : scp bin/hello remoteuser@remotehost:/remote/directory This example: Uses the golang:alpine container image.

Runs a job for building code. Stores build executable as artifact. Runs a second job to deploy to staging , which also: Requires the build job to succeed before running. Requires the commit target branch staging . Uses the build executable artifact. The equivalent GitLab CI/CD .gitlab-ci.yml file would be: yaml default : image : golang:alpine stages : - build - deploy build-job : stage : build script : - apk update - go build -o bin/hello artifacts : paths : - bin/hello expire_in : 1 week deploy-job : stage : deploy script : - echo "Deploying to Staging" - scp bin/hello remoteuser@remotehost:/remote/directory rules : - if : $CI_COMMIT_BRANCH == 'staging' Parallel In both GitHub and GitLab, Jobs run in parallel by default.

For example, in a GitHub Actions workflow file: yaml on : [ push] jobs : python-version : runs-on : ubuntu-latest container : python:latest steps : - run : python --version java-version : if : contains( github.ref, 'staging') runs-on : ubuntu-latest container : openjdk:latest steps : - run : java -version This example runs a Python job and a Java job in parallel, using different container images. The Java job only runs when the staging branch is changed.

The equivalent GitLab CI/CD .gitlab-ci.yml file would be: yaml python-version : image : python:latest script : - python --version java-version : image : openjdk:latest rules : - if : $CI_COMMIT_BRANCH == 'staging' script : - java -version In this case, no extra configuration is needed to make the jobs run in parallel. Jobs run in parallel by default, each on a different runner assuming there are enough runners for all the jobs.

The Java job is set to only run when the staging branch is changed. Matrix In both GitLab and GitHub you can use a matrix to run a job multiple times in parallel in a single pipeline, but with different variable values for each instance of the job. For example, in a GitHub Actions workflow file: yaml on : [ push] jobs : build : runs-on : ubuntu-latest steps : - run : echo "Building $PLATFORM for $ARCH" strategy : matrix : platform : [ linux, mac, windows] arch : [ x64, x86] test : runs-on : ubuntu-latest steps : - run : echo "Testing $PLATFORM for $ARCH" strategy : matrix : platform : [ linux, mac, windows] arch : [ x64, x86] deploy : runs-on : ubuntu-latest steps : - run : echo "Deploying $PLATFORM for $ARCH" strategy : matrix : platform : [ linux, mac, windows] arch : [ x64, x86] The equivalent GitLab CI/CD .gitlab-ci.yml file would be: yaml stages : - build - test - deploy .parallel-hidden-job : parallel : matrix : - PLATFORM : [ linux, mac, windows] ARCH : [ x64, x86] build-job : extends : .parallel-hidden-job stage : build script : - echo "Building $PLATFORM for $ARCH" test-job : extends : .parallel-hidden-job stage : test script : - echo "Testing $PLATFORM for $ARCH" deploy-job : extends : .parallel-hidden-job stage : deploy script : - echo "Deploying $PLATFORM for $ARCH" Trigger GitHub Actions requires you to add a trigger for your workflow.

GitLab is integrated tightly with Git, so SCM polling options for triggers are not needed, but can be configured per job if required. Sample GitHub Actions configuration: yaml on : push : branches : - main The equivalent GitLab CI/CD configuration would be: yaml rules : - if : '$CI_COMMIT_BRANCH == main' Pipelines can also be scheduled by using Cron syntax . Container images With GitLab you can run your CI/CD jobs in separate, isolated Docker containers by using the image keyword.

For example, in a GitHub Actions workflow f

The rest of this page is on the live site.