---

Steps to Implement

1. Create a Spring Boot Application

• Generate a Spring Boot project (e.g., using Spring Initializr).

2. Prepare Kubernetes Cluster

• Set up an AKS cluster in Azurw or EKS in AWS

3. Set Up Azure DevOps

• Create a service connection for the Docker registry.
• Create a service connection for the Kubernetes cluster.

4. Push Code to Azure Repos

• Add your application files (Dockerfile, deployment.yaml, and azure-pipelines.yml) to your repository.

5. Run the Pipeline

• The pipeline builds and pushes the Docker image, then deploys it to Kubernetes.

---

Key Updates

Readiness Probe

• Uses an HTTP GET request to check the /actuator/health endpoint, a common Spring Boot health check.
• Ensures the container is ready before receiving traffic.

Image Version Tag

• Replaced latest with <version-tag>. Replace <version-tag> with a specific version, e.g., v1.0.0.

Horizontal Pod Autoscaler (HPA)

• Added an HPA to scale the deployment based on CPU usage.
• Adjusts the replicas between 2 and 10 depending on average CPU utilization.

Liveness Probe

• Path: /actuator/dependency-health assumes your Spring Boot application exposes an endpoint that checks the health of the dependent service.
• If this endpoint indicates failure (e.g., by returning a non-2xx HTTP status), Kubernetes will restart the pod.
• Ensure your application has logic to query the dependent service and respond appropriately at this endpoint.

Initial Delay

• initialDelaySeconds: 15 ensures Kubernetes waits long enough for the dependent service to start before beginning the liveness checks.

Failure Handling

• failureThreshold: 3 restarts the pod only after three consecutive failures.

---

Instructions

1. Update your Docker image version during the build process in Azure DevOps to tag the image with a version instead of latest.

2. Deploy the updated deployment.yaml with your pipeline or manually using the command:

   kubectl apply -f deployment.yaml

3. Verify scaling behavior:

   • Use the command below to check HPA metrics:

     kubectl get hpa

   • Perform a load test on the application to observe pods scaling up or down.

---