KUBERNETES-README.md

Microservices Deployment Guide for Kubernetes

Overview

This guide details the deployment of our microservices (Order-Service, Inventory-Service, and Billing-Service) on Kubernetes using Minikube. The deployment includes:

• Three Spring Boot microservices
• MySQL database
• Apache Kafka
• Kubernetes Deployments, Services and ConfigMaps

Prerequisites

1. Minikube: installed and running

2. kubectl CLI tool installed (installed with minikube installation)

   • Verify with kubectl version

3. Docker installed (for minikube)

Environment Setup

1. Start Minikube

• Start minikube with minikube start and verify the status.

minikube start

minikube status

2. Enable Minikube Docker Environment

eval $(minikube docker-env)

This command configures your terminal to use Minikube's Docker daemon.

Building Docker Images

Build the Docker images for each service within Minikube's environment:

• Navigate to each service order-service,inventory-service and billing-service where DockerFile is located.

• Build docker images for all services one by one.

Docker image for zookeeper, kafka and mysql is pulled by kubectl through deployment.

# Build Order Service
cd order-service
docker build -t order-service:latest .

# Build Inventory Service
cd ../inventory-service
docker build -t inventory-service:latest .

# Build Billing Service
cd ../billing-service
docker build -t billing-service:latest .

Docker Build : Order Service

Docker Build : Inventory Service

Docker Build : Billing Service

• Verify docker images

  docker images

  

Deploying Infrastructure Components

Step 1:

Deploy database

To deploy MySQL with the provided YAML files, the resources need to be created in a specific order to ensure all dependencies are resolved properly. Here's the recommended order:

1. Create the Secret File: mysql-secret.yaml
   The MySQL deployment requires the root password from the secret. This must exist before the deployment starts.

   kubectl apply -f mysql-secret.yaml

2. Create the ConfigMap File: mysql-configmap.yaml The ConfigMap contains the initialization script for the database. This must be mounted as a volume when the MySQL container starts.

   kubectl apply -f mysql-configmap.yaml

3. Create the PersistentVolumeClaim (PVC) File: mysql-pvc.yaml The PVC provides persistent storage for MySQL's data. The deployment will mount this PVC, so it must be created first.

   kubectl apply -f mysql-pvc.yaml

4. Deploy MySQL Service File: mysql-service.yaml The service allows other pods in the Kubernetes cluster to communicate with the MySQL database by abstracting the pod IPs. Exposes MySQL on port 3306 (the default port for MySQL).

    kubectl apply -f mysql-service.yaml

5. Deploy MySQL File: mysql-deployment.yaml The deployment will reference the Secret, ConfigMap, and PVC. Ensure these resources are created beforehand.

    kubectl apply -f mysql-deployment.yaml

Run the commands in the following order:

kubectl apply -f mysql-secret.yaml
kubectl apply -f mysql-configmap.yaml
kubectl apply -f mysql-pvc.yaml
kubectl apply -f mysql-service.yaml
kubectl apply -f mysql-deployment.yaml

After applying these resources, you can verify the deployment:

# Check the pods
kubectl get pods

# Check the Persistent Volume and Persistent Volume Claim
kubectl get pv
kubectl get pvc

# Check the ConfigMap and Secret
kubectl get configmap
kubectl get secret

kubectl get deployments
kubectl get service

Step 2:

Deploy all other services

Navigate to infrastructure directory and run the following command. The following commands create necessary config, deployment,services for zookeeper, kafka, order-service, inventory-service and billing-service.

kubectl apply -f apply-all-kube-config.yaml

• kubectl pull the images for mysql, zookeeper, kafka.

Verify with following

```bash
docker images
```
```bash
docker ps
```
![kube11.png](screenshots%2Fkubernetes%2Fkube11.png)

Verify the deployment,pods,service with following command

```bash
kubectl get all
```

Once all pods are up and running, verify the following

• Kafka topics is created.
• All tables are created in mysql
• Service is able to produce message to topic.
• Service is able to consume the message.
• Verify the tables are updated

1.Kafka topics is created.

Execute the kafka pod with following command

kubectl exec -it <your-pod-name> -- /bin/bash

kubectl exec -it kafka-8cf75665c-57wt2 -- /bin/bash

Navigate to kafka directory.

cd /opt/kafka

Get the list of kafka-topics with following command:

./bin/kafka-topics.sh --bootstrap-server localhost:9092 --list

Get the list of consumer group id.

  ./bin/kafka-consumer-groups.sh --bootstrap-server localhost:9092 --list

billing-group-id and inventory-group-id are created and ready to consume the messages

2.All tables are created in mysql

Execute the mysql pod with following command

kubectl exec -it <your-pod-name> -- /bin/bash

kubectl exec -it mysql-55d56f565d-fckkb -- /bin/bash

Login to mysql server

mysql -u root -p

Enter password on prompt.

Verify the database with following

show databases;

use product_database;

show tables;

All tables are created.

3.Service is able to produce message to topic

The order-service app is used for producing message to the topic. It provides POST endpoint to produce message.

The endpoint can be executed within and outside the cluster.

• Within cluster

  To execute the endpoint with curl, execute order-service pod and use curl to hit the endpoint.

  • Use following command to execute order-service pod

  kubectl exec -it <your-pod-name> -- /bin/bash

   kubectl exec -it order-service-deployment-7797579499-f4hxv -- /bin/bash

  • Once executed inside pod, use curl command to hit endpoint

  curl -X POST http://localhost:8080/orders/place-order -H "Content-Type: application/json" -d '{"productName":"Samsung Galaxy S23","quantity":1,"price":999.99}'

  

  • Verify the application logs

  kubectl logs -f <pod_name>

  kubectl logs -f order-service-deployment-7797579499-f4hxv
  kubectl logs -f billing-service-deployment-5bcd779767-pbskn
  kubectl logs -f inventory-service-deployment-75bbbfcb9f-f8rfr

  

4.Verify message is consumed

• Execute the kafka pod with following command

kubectl exec -it <your-pod-name> -- /bin/bash

kubectl exec -it kafka-8cf75665c-57wt2 -- /bin/bash

• Check the messages consumed in order-topic
  • Navigate to kafka directory

  cd /opt/kafka  

  • Enter the following command

    ./bin/kafka-console-consumer.sh --bootstrap-server localhost:9092 --topic order-topic --from-beginning

  

5.Verify the tables are updated

Execute the mysql pod with following command

kubectl exec -it <your-pod-name> -- /bin/bash

kubectl exec -it mysql-55d56f565d-fckkb -- /bin/bash

Login to mysql server

mysql -u root -p

Enter password on prompt.

Verify the database with following.

show databases;

use product_database;

select * from inventory;

select * from orders;

select * from billing;

All tables are updated. Services are able to communicate with each other with kafka messaging.

Auto Scaling

HorizontalPodAutoscaler is the configuration for a horizontal pod autoscaler, which automatically manages the replica count of any resource implementing the scale subresource based on the metrics specified.

Find autoscaling.yaml file in infrastructure/autoscaling

kubectl apply -f autoscaling.yaml

kubectl get hpa

kubectl describe hpa <hpa-name>

• Troubleshooting auto scaler

  Cluster is missing the Metrics Server, which HPA needs to collect resource metrics.

  You might see the following error when you describe the hpa.

    Warning  FailedGetResourceMetric       12s (x3 over 42s)  horizontal-pod-autoscaler  failed to get cpu utilization: unable to get metrics for resource cpu: unable to fetch metrics from resource metrics API: the server is currently unable to handle the request (get pods.metrics.k8s.io)
    Warning  FailedGetResourceMetric       12s (x3 over 42s)  horizontal-pod-autoscaler  failed to get memory utilization: unable to get metrics for resource memory: unable to fetch metrics from resource metrics API: the server is currently unable to handle the request (get pods.metrics.k8s.io)

  Install metrics server with following command

  kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

  Enable in minikube with:

  minikube addons enable metrics-server

ToDo :

The high memory usage you're seeing is likely due to Java's default heap space allocation in Spring Boot applications. The JVM allocates a percentage of the container's memory limit (512Mi) at startup, even if the application isn't actively using it.

To optimize this, add these JVM arguments to your container specs:

env:
  - name: JAVA_OPTS
    value: "-Xms128m -Xmx256m -XX:+UseG1GC"
  - name: SPRING_HEAP_MEMORY
    value: "-XX:MaxRAMPercentage=50.0"

This will:

• Set initial heap size to 128MB
• Limit max heap to 256MB
• Use G1 garbage collector
• Limit JVM memory to 50% of container memory

Monitoring

Monitor your deployment using:

# Get pod status
kubectl get pods 

# Get pod logs
kubectl logs -f <pod-name>

# Get resource usage
kubectl top pods

<screenshot: Show monitoring dashboard or CLI output>

Cleanup

To remove all deployed resources:

# Delete all deployments from default namespace
kubectl delete all --all -n default

# Stop Minikube
minikube stop

Additional Notes

• All services are configured to use internal Kubernetes DNS for service discovery
• Persistent volumes are used for MySQL data storage
• Services are exposed through Kubernetes Services
• ConfigMaps are used for environment-specific configurations
• Secrets are used for sensitive data like passwords

For any issues or improvements, please create an issue in the repository.