Kubernetes and Argo Workflows Demo Project

This project demonstrates a simple CI/CD pipeline using Kubernetes (k8s) and Argo Workflows. It deploys a basic Node.js application to a Kubernetes cluster using Argo Workflows for orchestration.

Prerequisites

• Docker installed and configured
• Kubernetes CLI (kubectl) installed
• Minikube, Kind, or another local Kubernetes solution
• Git

Setup Instructions

1. Clone this repository

git clone <your-repo-url>
cd k8s-argo-demo

2. Set up a local Kubernetes cluster and Argo Workflows

Run the setup script to create a local Kubernetes cluster and install Argo Workflows:

chmod +x scripts/setup.sh
./scripts/setup.sh

This script:

• Starts a Minikube cluster
• Installs Argo Workflows
• Sets up port forwarding to access the Argo UI

3. Build and push the Docker image

Replace ${YOUR_DOCKER_USERNAME} in the Kubernetes manifests and Argo Workflow template with your Docker Hub username:

cd app
docker build -t yourusername/k8s-argo-demo:latest .
docker push yourusername/k8s-argo-demo:latest

4. Deploy the application manually (Optional)

Apply the Kubernetes manifests:

kubectl apply -f kubernetes/manifests.yaml

Check that your application is running:

kubectl get pods -n demo-app

5. Run the Argo Workflow

Submit the workflow to Argo:

kubectl apply -f argo/workflow.yaml

6. Access the application

kubectl port-forward -n demo-app svc/demo-app-service 8080:80

Then open http://localhost:8080 in your browser.

Project Structure

• app/: Contains the Node.js application code and Dockerfile
• kubernetes/: Contains Kubernetes deployment manifests
• argo/: Contains Argo Workflow templates
• scripts/: Contains setup and utility scripts

Learning Objectives Covered

• Setting up a local Kubernetes cluster
• Creating a basic containerized application
• Writing Kubernetes deployment manifests
• Understanding Kubernetes namespaces, deployments, services, and ingress
• Creating and running Argo Workflows
• Implementing a simple CI/CD pipeline

Next Steps

Once you're comfortable with this basic setup, you can:

1. Add automated testing to your workflow
2. Implement environment-specific deployments (dev, staging, prod)
3. Add Argo CD for GitOps-style continuous delivery
4. Explore Argo Events for event-driven workflows
5. Implement rollback mechanisms for failed deployments

Troubleshooting

• If you encounter issues with Docker socket permissions in Argo Workflows, make sure you've mounted the Docker socket correctly and have the necessary permissions.
• For Minikube, you might need to run eval $(minikube docker-env) to use the Minikube Docker daemon.
• If your ingress is not working, make sure you have an ingress controller installed in your cluster.