DevOps Project · Microsoft Azure

Multi-Environment CI/CD on Azure Kubernetes Service

An end-to-end GitHub Actions pipeline that tests, containerises and deploys a Node.js application to development and production environments in Azure Kubernetes Service.

View Live Deployment View Source Code Back to Projects

This link opens the live application when the Azure AKS cluster is active. The cluster is currently stopped to minimise hosting costs.

Project Overview

From source change to verified deployment

This project demonstrates a complete CI/CD workflow rather than only building a Docker image.

Feature-branch changes run automated tests and deploy to the development namespace. After the change is reviewed and merged into main, the workflow prepares the production release.

The production deployment uses a protected GitHub Environment. The workflow pauses until the release is manually approved before deploying into the production namespace.

Both environments run in Azure Kubernetes Service and use NGINX Ingress to route external traffic to the application.

Architecture

Application delivery flow

01

GitHub Source

Code is pushed to a feature branch and reviewed before merging into main.

02

Quality Gates

Unit tests run against Node.js 18, 19 and 20, alongside code coverage.

03

Container Image

Docker images are built and tagged using the Git commit SHA.

04

AKS Deployment

Feature branches deploy to development. Main deploys to production after approval.

05

Integration Test

GitHub Actions calls the externally routed /live endpoint.

Pipeline Design

Automated workflow stages

Matrix Unit Testing

Executes automated tests across Node.js versions 18, 19 and 20.

Code Coverage

Measures application test coverage and publishes workflow artifacts.

Containerisation

Builds Docker images and publishes versioned images to container registries.

Development Deployment

Feature-branch changes deploy automatically to the development namespace.

Production Approval

The protected production environment requires manual approval before deployment.

Integration Testing

Verifies the deployed application through its externally accessible health endpoint.

Release Strategy

Development and production separation

Source Event Target Control Validation
Push to feature/* development namespace Automatic after CI succeeds Development /live test
Merge or push to main production namespace Manual GitHub Environment approval Production /live test

Protected production deployment

A successful main-branch build does not deploy immediately. The production job pauses until the deployment is reviewed and manually approved.

Kubernetes Implementation

AKS resources

Deployments

Manage application replicas and rolling updates in development and production.

Services

ClusterIP Services expose the Node.js application internally on port 3000.

NGINX Ingress

Routes external traffic from the Azure Load Balancer to the correct Kubernetes Service.

Repository structure 

.github/workflows/solar-system.yml
Dockerfile
kubernetes/
├── development/
│   ├── deployment.yaml
│   ├── service.yaml
│   └── ingress.yaml
└── production/
    ├── deployment.yaml
    ├── service.yaml
    └── ingress.yaml

Deployment Evidence

Pipeline and AKS results

Permanent evidence from the completed CI/CD pipeline and Azure Kubernetes Service deployment.

Development CI/CD Pipeline

Successful GitHub Actions development pipeline

A feature-branch workflow completing matrix unit testing, code coverage, containerisation, development deployment and integration testing. The production jobs remain skipped because the workflow was triggered from a feature branch.

Approved Production Deployment

Successful production deployment after manual approval

The main-branch workflow after its protected production environment was manually approved. The production deployment and post-deployment integration test both completed successfully.

Healthy AKS Workloads

Healthy development and production deployments in AKS

Azure AKS running the Solar System application in separate development and production namespaces. The NGINX Ingress Controller and supporting Kubernetes services are also healthy.

Deployed Application

Solar System application deployed to an AKS pod

The containerised Solar System application served through the production NGINX Ingress. The pod name displayed by the application confirms that the response was generated by a Kubernetes pod.

Commit-to-Deployment Verification

Commit-tagged Docker image and running production pods in Azure AKS

The production Kubernetes Deployment uses a Docker image tagged with the Git commit SHA. The same image is running across two healthy production pods in Azure AKS. This provides traceability from the GitHub commit and workflow run to the container image deployed in Kubernetes.

Security Controls

Securing the delivery process

Security controls protect application credentials, separate deployment environments and prevent unapproved production releases.

  • Sensitive MongoDB passwords are stored as GitHub Secrets rather than committed to source control.
  • Kubernetes Secrets provide application credentials to the running containers.
  • Production deployment uses a protected GitHub Environment with manual approval.
  • Development and production workloads are separated using Kubernetes namespaces.
  • Docker images use immutable commit-based tags for deployment traceability.
  • GitHub Actions runs automated tests before any environment deployment begins.

Project Outcome

Skills demonstrated

The completed project connects source control, testing, container delivery, Kubernetes deployment and operational verification through a controlled release process.

  • Branch-aware multi-environment deployments
  • Azure Kubernetes Service administration
  • Docker image creation and SHA-based tagging
  • Automated unit and integration testing
  • Manual production release approval
  • Kubernetes Ingress and external routing
View GitHub Repository Back to Projects