CI/CD Pipelines¶

CI/CD is the automation backbone of modern software delivery. Continuous Integration (CI) builds and tests every change as soon as it is pushed. Continuous Delivery/Deployment (CD) packages that change — usually as a container image — and ships it to an environment automatically.

This section is a practical, tool-by-tool deep dive. Every page includes installation and setup guidance, real YAML you can copy, and diagrams of how the pieces connect.

What you will learn in this section

How a commit flows through test → build → image → registry → deploy
GitHub Actions from zero to production pipelines
Complete pipelines for a Java (Spring Boot) and a Python (FastAPI) web app
GitLab CI/CD with runners and the built-in container registry
GitOps with ArgoCD — letting a controller deploy for you
Where Jenkins fits and how its model compares

The Big Picture¶

Every CI/CD pipeline, regardless of tool, is a variation of this flow:

flowchart LR
    DEV[Developer<br>git push] --> CI[CI Pipeline<br>lint + unit tests]
    CI --> BUILD[Build Artifact<br>JAR / wheel]
    BUILD --> IMG[Docker Image<br>multi-stage build]
    IMG --> REG[(Container Registry<br>GHCR / Docker Hub / GitLab)]
    REG --> CD[CD Stage<br>deploy job or GitOps]
    CD --> ENV[Environment<br>VM / Kubernetes / Cloud]
    classDef dev fill:#fef3c7,stroke:#d97706,color:#78350f
    classDef ci fill:#dbeafe,stroke:#2563eb,color:#172554
    classDef img fill:#ede9fe,stroke:#7c3aed,color:#3b0764
    classDef deploy fill:#dcfce7,stroke:#16a34a,color:#14532d
    class DEV dev
    class CI,BUILD ci
    class IMG,REG img
    class CD,ENV deploy

The stages map to concrete tool features:

Stage	GitHub Actions	GitLab CI	Jenkins	ArgoCD
Trigger	`on: push`	`rules:`	webhook / poll	Git watch (GitOps)
Test	job + steps	`stage: test`	`stage('Test')`	not its job
Build image	`docker/build-push-action`	`docker build` in job	`docker.build()`	not its job
Store image	GHCR / Docker Hub	GitLab Registry	any registry	reads registry via manifests
Deploy	deploy job / environment	`environment:`	`stage('Deploy')`	automatic sync

CI tools push, ArgoCD pulls

GitHub Actions, GitLab CI, and Jenkins push changes to servers. ArgoCD flips the model: it runs inside the cluster and pulls the desired state from Git. Most production setups combine both — CI builds the image, ArgoCD deploys it.

Choosing a Tool¶

flowchart TD
    Q1{Where is<br>your code?} -->|GitHub| GHA[GitHub Actions]
    Q1 -->|GitLab| GL[GitLab CI/CD]
    Q1 -->|Self-hosted /<br>legacy needs| JK[Jenkins]
    GHA --> Q2{Deploying to<br>Kubernetes?}
    GL --> Q2
    JK --> Q2
    Q2 -->|Yes| ARGO[Add ArgoCD<br>for GitOps CD]
    Q2 -->|No| DIRECT[Deploy from<br>the pipeline]
    classDef q fill:#fef3c7,stroke:#d97706,color:#78350f
    classDef tool fill:#dbeafe,stroke:#2563eb,color:#172554
    classDef gitops fill:#dcfce7,stroke:#16a34a,color:#14532d
    class Q1,Q2 q
    class GHA,GL,JK,DIRECT tool
    class ARGO gitops

GitHub Actions — zero infrastructure to start, huge marketplace of reusable actions, free minutes for public repos. The default choice if your code lives on GitHub.
GitLab CI/CD — a single application for repos, pipelines, and a container registry. Excellent for self-hosted setups.
Jenkins — self-hosted, plugin-driven, infinitely customizable. Common in enterprises with existing investment. See the Jenkins guide.
ArgoCD — not a CI tool at all; a Kubernetes GitOps controller that continuously reconciles your cluster with Git.

Section Contents¶

Page	What it covers
GitHub Actions Deep Dive	Concepts, workflow anatomy, runners, secrets, caching, matrix builds, self-hosted runner setup
Java Pipeline: Code to Production	Spring Boot app — JUnit tests, Maven build, multi-stage Docker image, push to GHCR, deploy
Python Pipeline: Code to Production	FastAPI app — pytest, ruff, Docker image, push, deploy
GitLab CI/CD	`.gitlab-ci.yml`, runner installation, built-in registry, Java and Python examples
ArgoCD and GitOps	Install ArgoCD on Kubernetes, Application manifests, CI + GitOps end-to-end flow
Jenkins	Installation on Ubuntu, first pipeline, core concepts