GitLab CI/CD: Pipelines, Runners & Auto DevOps Guide

In this tutorial, you'll learn about GitLab CI/CD: Pipelines, Runners & Auto DevOps Guide. We cover key concepts, practical examples, and best practices.

GitLab CI/CD pipelines automate builds, tests, and deployments using a .gitlab-ci.yml configuration file with stages that run jobs in sequence.

In this tutorial, you'll learn GitLab CI/CD for automating builds, tests, and deployments. GitLab CI/CD is built directly into GitLab, providing a seamless integration between your code repository and pipelines. By the end, you'll write .gitlab-ci.yml files, configure runners, use environments, and implement deployment strategies.

flowchart TD
  A[Git Push] --> B[Pipeline Triggered]
  B --> C[Stage: build]
  C --> D[Stage: test]
  D --> E[Stage: deploy]
  E --> F{Environment}
  F --> G[Staging]
  F --> H[Production]

Writing Your First Pipeline

Create .gitlab-ci.yml in your repository root:

stages:
  - build
  - test
  - deploy

build-job:
  stage: build
  script:
    - echo "Compiling application..."
    - mkdir -p build
    - echo "Build complete" > build/output.txt
  artifacts:
    paths:
      - build/

test-job:
  stage: test
  script:
    - echo "Running tests..."
    - test -f build/output.txt
    - echo "Tests passed"

deploy-job:
  stage: deploy
  script:
    - echo "Deploying to production..."
  only:
    - main

Expected output on commit: A pipeline runs with three sequential stages. Each stage shows its output in the GitLab UI.

Defining Jobs with Docker

Use Docker images for consistent environments:

test:
  image: node:18-alpine
  stage: test
  script:
    - npm ci
    - npm test
  cache:
    key: $CI_COMMIT_REF_SLUG
    paths:
      - node_modules/

GitLab Runners

Runners execute jobs. You can use shared GitLab-hosted runners or install your own:

# Register a self-managed runner
sudo gitlab-runner register \
  --url https://gitlab.com/ \
  --registration-token YOUR_TOKEN \
  --executor docker \
  --description "Docker Runner" \
  --docker-image alpine:latest

Environments and Deployments

Define deployment environments:

deploy-staging:
  stage: deploy
  environment:
    name: staging
    url: https://staging.example.com
  script:
    - echo "Deploy to staging"
  only:
    - develop

deploy-production:
  stage: deploy
  environment:
    name: production
    url: https://example.com
  script:
    - echo "Deploy to production"
  when: manual
  only:
    - main

Auto DevOps

Enable Auto DevOps in your project settings for automated CI/CD without writing a pipeline. GitLab automatically detects your language, runs tests, builds a Docker image, and deploys to Kubernetes.

Pipeline Components

CI/CD Variables and Secrets

Store sensitive data in GitLab CI/CD variables:

# Access in jobs
deploy:
  script:
    - echo "$DEPLOY_KEY" > deploy_key.pem
    - chmod 600 deploy_key.pem
    - ssh -i deploy_key.pem user@server "deploy"

Set variables in Project > Settings > CI/CD > Variables. Masked variables are hidden in job logs.

Child Pipelines and Multi-Project Pipelines

Compose pipelines from multiple projects:

# Trigger a child pipeline from another project
trigger-deploy:
  stage: deploy
  trigger:
    project: my-group/deployment-tool
    branch: main
    strategy: depend

Child pipelines let you build, test, and deploy across multiple projects in sequence.

GitLab CI/CD Templates

GitLab provides ready-made pipeline templates:

# Include a template
include:
  - template: Auto-DevOps.gitlab-ci.yml
  - template: Security/SAST.gitlab-ci.yml
  - template: Security/Secret-Detection.gitlab-ci.yml
  - project: 'my-group/my-shared-config'
    file: '/templates/ci.yml'

Common Errors

Practice Questions

Challenge

Create a .gitlab-ci.yml for a Python application with four stages: build (install dependencies), test (pytest with coverage), lint (flake8), and deploy (to staging and production). Configure caching for pip packages, use Docker images, and set up manual approval for production deployment. Include branch restrictions — staging from develop, production from main.

Real-World Task

Set up a GitLab project with Auto DevOps enabled. Push a sample application, trigger the pipeline, and verify each stage runs successfully. Then install a self-managed GitLab runner on a local machine using Docker executor. Configure it to run your pipelines without using shared runner minutes. This setup is used at DodaTech for internal tools where data security requires self-hosted runners.

Built by the developers of Doda Browser, DodaZIP, and Durga Antivirus Pro.