Skip to main content

Cloud Setup with EKS: AWS Luxury Hotel

EKS is AWS's managed Kubernetes. Like a luxury hotel. Managed. High availability. Production-ready. That's EKS.

🎯 The Big Picture​

Think of EKS like a luxury hotel managed by AWS. You don't manage the hotel. AWS does. High availability. Production-ready. That's EKS.

EKS (Elastic Kubernetes Service) is AWS's managed Kubernetes. Control plane managed by AWS. You manage worker nodes. Production-ready.

Why EKS?​

Perfect for:

  • Production workloads
  • AWS ecosystem
  • High availability
  • Enterprise needs

Why:

  • Managed control plane
  • AWS integration
  • High availability
  • Production-ready

Think of it as: AWS luxury hotel. Managed. High availability. Production.

Prerequisites​

What you need:

  • AWS account
  • AWS CLI installed
  • kubectl installed
  • eksctl (recommended) or AWS Console

Install eksctl:

# macOS
brew install eksctl

# Linux
curl --silent --location "https://github.com/weaveworks/eksctl/releases/latest/download/eksctl_$(uname -s)_amd64.tar.gz" | tar xz -C /tmp
sudo mv /tmp/eksctl /usr/local/bin

Verify:

eksctl version

Creating EKS Cluster​

Create cluster:

eksctl create cluster \
  --name my-hotel-cluster \
  --region us-east-1 \
  --nodegroup-name workers \
  --node-type t3.medium \
  --nodes 3 \
  --nodes-min 2 \
  --nodes-max 5

What happens:

  • Creates VPC
  • Creates control plane
  • Creates worker nodes
  • Configures kubectl

Takes 15-20 minutes. Be patient.

Using AWS Console​

Step 1: Create cluster

  • Go to EKS console
  • Click "Create cluster"
  • Configure cluster
  • Create

Step 2: Create node group

  • Go to cluster
  • Add node group
  • Configure nodes
  • Create

Step 3: Configure kubectl

aws eks update-kubeconfig --name my-hotel-cluster --region us-east-1

The AWS Luxury Hotel Analogy​

Think of EKS like AWS luxury hotel:

AWS manages: Hotel management (control plane)

  • You don't manage it
  • AWS handles it
  • High availability

You manage: Hotel floors (worker nodes)

  • You configure them
  • You scale them
  • You maintain them

Integration: Hotel services (AWS services)

  • IAM for security
  • VPC for networking
  • CloudWatch for monitoring

Once you see it this way, EKS makes perfect sense.

Verifying Setup​

Check cluster:

kubectl get nodes

Expected output:

NAME                                           STATUS   ROLES    AGE   VERSION
ip-192-168-1-100.us-east-1.compute.internal   Ready    <none>   5m    v1.28.0
ip-192-168-1-101.us-east-1.compute.internal   Ready    <none>   5m    v1.28.0
ip-192-168-1-102.us-east-1.compute.internal   Ready    <none>   5m    v1.28.0

Check cluster info:

kubectl cluster-info

Check AWS integration:

kubectl get configmap aws-auth -n kube-system

That's verification. Everything working.

Node Groups​

What they are:

  • Groups of worker nodes
  • Same instance type
  • Same configuration
  • Auto-scaling

Create node group:

eksctl create nodegroup \
  --cluster=my-hotel-cluster \
  --name=workers \
  --node-type=t3.large \
  --nodes=3 \
  --nodes-min=2 \
  --nodes-max=10

Think of it as: Hotel floor types. Standard rooms. Suite rooms. Different configurations.

Auto-Scaling​

Enable cluster autoscaler:

# Install cluster autoscaler
kubectl apply -f https://raw.githubusercontent.com/kubernetes/autoscaler/master/cluster-autoscaler/cloudprovider/aws/examples/cluster-autoscaler-autodiscover.yaml

# Annotate deployment
kubectl annotate deployment cluster-autoscaler \
  -n kube-system \
  cluster-autoscaler.kubernetes.io/safe-to-evict="false"

Think of it as: Automatic room capacity. Scales up when busy. Scales down when quiet.

Real-World Example: Production EKS Setup​

Complete production setup:

Step 1: Create cluster:

eksctl create cluster \
  --name production-hotel \
  --region us-east-1 \
  --version 1.28 \
  --nodegroup-name workers \
  --node-type t3.large \
  --nodes 3 \
  --nodes-min 2 \
  --nodes-max 10 \
  --managed

Step 2: Configure kubectl:

aws eks update-kubeconfig --name production-hotel --region us-east-1

Step 3: Verify:

kubectl get nodes
kubectl cluster-info

Step 4: Deploy application:

kubectl create deployment my-app --image=my-app:1.0
kubectl expose deployment my-app --port=80 --type=LoadBalancer

That's production EKS. Complete. Ready.

Cost Considerations​

EKS costs:

  • Control plane: $0.10/hour (~$73/month)
  • Worker nodes: EC2 instance costs
  • Data transfer: AWS data transfer costs

Optimization:

  • Use spot instances
  • Right-size nodes
  • Use Fargate (serverless)
  • Monitor costs

Think of it as: Hotel costs. Management fee. Room costs. Services costs.

My Take: EKS Strategy​

Here's what I do:

Production:

  • Use EKS
  • Managed control plane
  • Auto-scaling nodes
  • High availability

Development:

  • Use local (minikube/kind)
  • Or EKS dev cluster
  • Cost-effective

The key: Use EKS for production. Local for development. Right tool for job.

Memory Tip: The AWS Luxury Hotel Analogy​

EKS = AWS luxury hotel

AWS manages: Hotel management You manage: Hotel floors Integration: Hotel services

Once you see it this way, EKS makes perfect sense.

Common Mistakes​

  1. Not configuring kubectl: Can't access cluster
  2. Wrong region: Cluster in wrong region
  3. No node groups: No worker nodes
  4. Not monitoring costs: Unexpected bills
  5. No auto-scaling: Manual scaling

Key Takeaways​

  1. EKS is managed Kubernetes - AWS manages control plane
  2. You manage worker nodes - Configure and scale
  3. High availability - Production-ready
  4. AWS integration - IAM, VPC, CloudWatch
  5. Cost considerations - Control plane + nodes

What's Next?​

Congratulations! You've completed the Setting Up Cluster module. Now let's learn about pods. Next: Understanding Pods.

Remember: EKS is like AWS luxury hotel. Managed control plane. You manage nodes. High availability. Production-ready. Right for production workloads.