# Check Kubernetes cluster
kubectl cluster-info
kubectl get nodes# Install metrics-server for resource monitoring
kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml
# Patch for Docker Desktop compatibility
kubectl patch deployment metrics-server -n kube-system --type='json' -p='[{"op": "add", "path": "/spec/template/spec/containers/0/args/-", "value": "--kubelet-insecure-tls"}]'
# Wait for metrics-server to be ready
kubectl wait --namespace kube-system --for=condition=ready pod --selector=k8s-app=metrics-server --timeout=60s
# Verify metrics are working
kubectl top nodes
kubectl top pods --all-namespaces# Install NGINX Ingress Controller (one-time setup)
kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v1.8.2/deploy/static/provider/cloud/deploy.yaml
# Wait for ingress controller to be ready
kubectl wait --namespace ingress-nginx --for=condition=ready pod --selector=app.kubernetes.io/component=controller --timeout=90s
# Add local DNS entries for all environments
echo "127.0.0.1 health-service-dev.local" | sudo tee -a /etc/hosts
echo "127.0.0.1 health-service-staging.local" | sudo tee -a /etc/hosts
echo "127.0.0.1 health-service-prod.local" | sudo tee -a /etc/hosts
# Test all environments
curl http://health-service-dev.local/health
curl http://health-service-staging.local/health
curl http://health-service-prod.local/health# Build the Go application image
docker build -f devops/docker/Dockerfile -t health-service:v1.0 .
# Verify image was created
docker images | grep health-service
# Test the image locally (optional)
docker run -p 8080:8080 health-service:v2.1
# Test: curl http://localhost:8080/health
# Stop: Ctrl+CMake sure the image tag in your values files matches the built image:
# In devops/helm/environments/values.dev.yaml
image_name: health-service
image_tag: v1.0 # ← Tag
# In devops/helm/environments/values.nonprod.yaml
image_tag: v1.0 # ← Tag
# In devops/helm/environments/values.prod.yaml
image_tag: v1.0 # ← TagNote: Since we use imagePullPolicy: Never for local development, Kubernetes will use the local Docker image you just built.
# Test deployment (dry-run)
helm upgrade --install health-service-dev devops/helm/ --namespace health-service-dev --create-namespace --values devops/helm/environments/values.dev.yaml --dry-run --debug
# Deploy development environment
helm upgrade --install health-service-dev devops/helm/ --namespace health-service-dev --create-namespace --values devops/helm/environments/values.dev.yaml
# Check deployment
kubectl get pods -n health-service-dev
kubectl get services -n health-service-dev
kubectl get ingress -n health-service-dev
kubectl get hpa -n health-service-dev# Deploy staging environment
helm upgrade --install health-service-staging devops/helm/ --namespace health-service-staging --create-namespace --values devops/helm/environments/values.nonprod.yaml
# Check deployment
kubectl get pods -n health-service-staging
kubectl get services -n health-service-staging
kubectl get ingress -n health-service-staging
kubectl get hpa -n health-service-staging# Deploy production environment
helm upgrade --install health-service-prod devops/helm/ --namespace health-service-prod --create-namespace --values devops/helm/environments/values.prod.yaml
# Check deployment
kubectl get pods -n health-service-prod
kubectl get services -n health-service-prod
kubectl get ingress -n health-service-prod
kubectl get hpa -n health-service-prod# Development
kubectl get all -n health-service-dev
kubectl get ingress -n health-service-dev
kubectl logs -n health-service-dev -l service=health-service-dev
# Staging
kubectl get all -n health-service-staging
kubectl get ingress -n health-service-staging
kubectl logs -n health-service-staging -l service=health-service-staging
# Production
kubectl get all -n health-service-prod
kubectl get ingress -n health-service-prod
kubectl logs -n health-service-prod -l service=health-service-prod# Remove specific environment
helm uninstall health-service-dev --namespace health-service-dev
helm uninstall health-service-staging --namespace health-service-staging
helm uninstall health-service-prod --namespace health-service-prod
# Remove all environments at once
kubectl delete namespace health-service-dev health-service-staging health-service-prod# Remove all health-service deployments
kubectl delete namespace health-service-dev health-service-staging health-service-prod
# Remove NGINX Ingress Controller
kubectl delete -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v1.8.2/deploy/static/provider/cloud/deploy.yaml
# Remove metrics-server (if you want to clean everything)
kubectl delete -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml
# Clean up local DNS entries (optional)
sudo sed -i '' '/health-service.*\.local/d' /etc/hosts
# Verify cleanup
kubectl get all --all-namespaces | grep -E "(health-service|ingress-nginx|metrics-server)"# Validate Helm chart (requires environment values)
helm lint devops/helm/ --values devops/helm/environments/values.dev.yaml
# Preview templates for specific environment
helm template health-service-dev devops/helm/ --values devops/helm/environments/values.dev.yaml
helm template health-service-staging devops/helm/ --values devops/helm/environments/values.nonprod.yaml
helm template health-service-prod devops/helm/ --values devops/helm/environments/values.prod.yaml# Check pod details (replace ENVIRONMENT with dev/staging/prod)
kubectl describe pod -n health-service-ENVIRONMENT -l service=health-service-ENVIRONMENT
# Check service endpoints
kubectl get endpoints -n health-service-ENVIRONMENT
# Validate ingress configuration
kubectl describe ingress -n health-service-ENVIRONMENT
# Check pod events for issues
kubectl get events -n health-service-ENVIRONMENT --sort-by='.lastTimestamp'# Check if ingress controller is running
kubectl get pods -n ingress-nginx
# Test internal cluster connectivity
kubectl run test-pod --rm -i --tty --image=busybox -- /bin/sh
# Inside pod, test each environment:
# wget -qO- http://health-service-dev-service.health-service-dev.svc.cluster.local/health
# wget -qO- http://health-service-staging-service.health-service-staging.svc.cluster.local/health
# wget -qO- http://health-service-prod-service.health-service-prod.svc.cluster.local/health
# View all resources across environments
kubectl get pods,services,ingress --all-namespaces | grep health-service| Environment | Min Replicas | Max Replicas | CPU Target | Memory Target | Scale Up Policy | Scale Down Policy |
|---|---|---|---|---|---|---|
| Development | 1 | 3 | 70% | 80% | Conservative (50%/1 pod per 60s) | Slow (10% per 60s) |
| Staging | 2 | 5 | 60% | 70% | Moderate (100%/2 pods per 30s) | Medium (20% per 60s) |
| Production | 3 | 10 | 50% | 60% | Aggressive (100%/3 pods per 15s) | Careful (25% per 60s) |
# Check HPA status across all environments
kubectl get hpa --all-namespaces | grep health-service
# Watch HPA scaling in real-time
kubectl get hpa -n health-service-dev -w
# Detailed HPA information
kubectl describe hpa -n health-service-dev health-service-dev-hpa
# Check current resource usage
kubectl top pods --all-namespaces | grep health-service# Generate sustained CPU load to trigger scaling
curl http://health-service-dev.local/load/30000 & # 30-second CPU load
curl http://health-service-dev.local/load/30000 & # 30-second CPU load
curl http://health-service-dev.local/load/30000 & # 30-second CPU load
# Test multiple concurrent requests to see load distribution
for i in {1..6}; do
curl http://health-service-dev.local/load/20000 &
done
# Watch scaling in real-time
kubectl get pods -n health-service-dev -w
kubectl get hpa -n health-service-dev -w
# Monitor which pods handle requests
kubectl logs -n health-service-dev -l service=health-service-dev --tail=20 -f# Install Apache Bench for load testing
# macOS: brew install httpd
# Ubuntu: sudo apt install apache2-utils
# Generate load to trigger scaling
ab -n 10000 -c 50 http://health-service-dev.local/health
# Monitor HPA decisions
kubectl get events -n health-service-dev --sort-by='.lastTimestamp' | grep HorizontalPodAutoscaler# Check if metrics-server is running
kubectl get pods -n kube-system | grep metrics-server
kubectl top nodes
# Verify pod resource requests (required for HPA)
kubectl describe pod -n health-service-dev -l service=health-service-dev
# Check HPA events and scaling decisions
kubectl describe hpa -n health-service-dev
# View HPA controller logs
kubectl logs -n kube-system -l k8s-app=metrics-server- Development: 1-3 replicas, minimal resources, conservative HPA
- Staging: 2-5 replicas, production-like resources, moderate HPA
- Production: 3-10 replicas, maximum resources, aggressive HPA