🚀 CI/CD Pipeline: Java → Docker → ECR → EKS

A production-style CI/CD pipeline that automates the deployment of a Java application from source code to a live environment on Amazon EKS.

---

📌 Overview

End-to-end pipeline that:

• Builds a Java (Spring Boot) application with Maven.
• Containerizes it using Docker
• Pushes the image to Amazon ECR
• Deploys to Amazon EKS using Kubernetes manifests
• Note: I replaced the real java app by a Java baseline since I can't share the real Java app for security reasons.

---

⚙️ Pipeline Stages

1. Source Checkout

• Pulls jenkins-jobs branch
• Uses Jenkins credentials for authentication

2. Build

• Runs mvn clean package
• Executes unit tests
• Produces JAR artifact

3. Docker Build & Push

• Builds image using Amazon Corretto base image
• Tags and pushes to ECR
• Authenticates with AWS

4. Deploy to EKS

• Configure kubeconfig:

  aws eks update-kubeconfig --region us-east-1 --name demo-cluster

• Deploy to cluster:

  kubectl apply -f kubernetes/

• Performs rolling updates

---

🧰 Tech Stack

• Jenkins (CI/CD) (Note: I used Jenkins because it was the default tool used by the workplace, however I would've used GitHub Actions or GitLab CI instead)
• Maven (Build)
• Docker (Containerization)
• AWS ECR (Registry)
• Kubernetes / EKS (Orchestration)
• AWS (Cloud)
• aws-cli, kubectl
• GitLab

---

⚠️ Challenges & Fixes

• AWS CLI missing in Jenkins → Installed CLI in execution environment
• kubectl could not connect → Generated kubeconfig via AWS CLI
• Wrong EKS cluster name → Verified directly in AWS Console
• Jenkins agent vs controller confusion → Ensured tools run in agent
• Docker permission issues → Fixed execution context and permissions

Pipeline required 23+ runs to stabilize (see pipelines_failures/)

---

✅ Result

• ✔ Application builds successfully
• ✔ Image pushed to ECR
• ✔ Kubernetes resources created
• ✔ Application deployed to EKS

---

💡 Key Learnings

• CI/CD failures are often environment-related, not code-related
• Strong understanding of Jenkins execution model (agent vs controller)
• Hands-on experience with:
  • AWS CLI in pipelines
  • ECR authentication
  • EKS kubeconfig setup
  • Kubernetes deployments

---

🧪 Takeaway

Go to evidence folder on project root to find screenshots. This project demonstrates the ability to design, implement, and debug a real-world CI/CD pipeline across Jenkins, Docker, AWS, and Kubernetes — focusing on reliability, not just implementation.