Amazon EKS — eksctl create cluster

3 minute read

eksctl create cluster

Let’s create your very first EKS cluster. One command, about 15 minutes, and you’ll have a fully functional Kubernetes cluster on AWS.

The Simplest Way to Create a Cluster

$ eksctl create cluster \
    --name demo \
    --region us-east-1 \
    --nodegroup-name demo-nodes \
    --node-type t3.medium \
    --nodes 2 \
    --nodes-min 1 \
    --nodes-max 3

Expected output (takes ~15 minutes):

2024-01-15 10:00:00 [ℹ]  eksctl version 0.175.0
2024-01-15 10:00:00 [ℹ]  using region us-east-1
2024-01-15 10:00:01 [ℹ]  subnets for us-east-1a - public:192.168.0.0/19 private:192.168.64.0/19
2024-01-15 10:00:01 [ℹ]  subnets for us-east-1b - public:192.168.32.0/19 private:192.168.96.0/19
2024-01-15 10:00:01 [ℹ]  nodegroup "demo-nodes" will use "" [AmazonLinux2/1.29]
2024-01-15 10:00:02 [ℹ]  using Kubernetes version 1.29
2024-01-15 10:00:02 [ℹ]  creating EKS cluster "demo" in "us-east-1" region
2024-01-15 10:00:02 [ℹ]  will create 2 separate CloudFormation stacks for cluster itself and the initial managed nodegroup
...
2024-01-15 10:14:30 [ℹ]  node "ip-192-168-75-112.ec2.internal" is ready
2024-01-15 10:14:30 [ℹ]  node "ip-192-168-98-43.ec2.internal" is ready
2024-01-15 10:14:32 [✔]  EKS cluster "demo" in "us-east-1" region is ready

What Just Happened?

That single command created over 30 AWS resources:

eksctl create cluster ──▶ CloudFormation Stacks
                              │
                              ├── eksctl-demo-cluster
                              │   ├── VPC
                              │   ├── 4 Subnets (2 public, 2 private)
                              │   ├── Internet Gateway
                              │   ├── 2 NAT Gateways
                              │   ├── Route Tables
                              │   ├── EKS Cluster
                              │   ├── IAM Cluster Role
                              │   ├── Security Groups
                              │   └── OIDC Provider
                              │
                              └── eksctl-demo-nodegroup-demo-nodes
                                  ├── IAM Node Role
                                  ├── Launch Template
                                  ├── Auto Scaling Group
                                  └── 2 EC2 Instances (t3.medium)

Understanding the eksctl Flags

Flag Purpose Default
--name Cluster name Auto-generated
--region AWS region Your CLI default region
--nodegroup-name Name for the node group Auto-generated
--node-type EC2 instance type m5.large
--nodes Number of worker nodes 2
--nodes-min Minimum nodes (for auto-scaling) Same as --nodes
--nodes-max Maximum nodes (for auto-scaling) Same as --nodes

Verify the Cluster Was Created

# Check cluster exists in AWS
$ aws eks describe-cluster --name demo --query 'cluster.status'
"ACTIVE"

# Check nodes are ready
$ kubectl get nodes
NAME                              STATUS   ROLES    AGE   VERSION
ip-192-168-75-112.ec2.internal   Ready    <none>   5m    v1.29.0-eks-...
ip-192-168-98-43.ec2.internal    Ready    <none>   5m    v1.29.0-eks-...

# Check system pods
$ kubectl get pods -n kube-system
NAME                       READY   STATUS    RESTARTS   AGE
aws-node-xxxxx             2/2     Running   0          5m
coredns-xxxxxxxxxx-xxxxx   1/1     Running   0          10m
coredns-xxxxxxxxxx-xxxxx   1/1     Running   0          10m
kube-proxy-xxxxx           1/1     Running   0          5m
kube-proxy-xxxxx           1/1     Running   0          5m

Deploy a Quick Test App

Let’s prove the cluster works:

# Deploy nginx
$ kubectl create deployment nginx --image=nginx --replicas=2

# Check pods
$ kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
nginx-7854ff8877-abcde   1/1     Running   0          30s
nginx-7854ff8877-fghij   1/1     Running   0          30s

# Expose it
$ kubectl expose deployment nginx --port=80 --type=LoadBalancer

# Wait for external IP (takes ~2 minutes)
$ kubectl get svc nginx -w
NAME    TYPE           CLUSTER-IP      EXTERNAL-IP                                                   PORT(S)        AGE
nginx   LoadBalancer   10.100.200.10   a1234567890abcdef-1234567890.us-east-1.elb.amazonaws.com      80:31234/TCP   2m

# Test it
$ curl a1234567890abcdef-1234567890.us-east-1.elb.amazonaws.com
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
...

Clean up the test app:

$ kubectl delete service nginx
$ kubectl delete deployment nginx

Common Mistakes

Mistake Impact Fix
Not specifying --region Cluster created in wrong region Always specify --region
Using m5.large (default) More expensive than needed for learning Use --node-type t3.medium
Forgetting to delete cluster Ongoing charges (~$170/month) eksctl delete cluster --name demo
Running as root AWS account Security risk Use IAM user/role

Quick Reference

# Create cluster
$ eksctl create cluster --name demo --region us-east-1 --node-type t3.medium --nodes 2

# Check status
$ eksctl get cluster --name demo --region us-east-1

# List node groups
$ eksctl get nodegroup --cluster demo --region us-east-1

# Delete cluster (IMPORTANT!)
$ eksctl delete cluster --name demo --region us-east-1

What’s Next?

The CLI approach is great for quick clusters, but production teams use config files. Let’s learn about Cluster Config YAML.

You may also enjoy