Create an Amazon EKS Cluster

Create an EKS Cluster

Create an EKS cluster with managed worker nodes and OIDC enabled:

eksctl create cluster \
  --name ${CLUSTER_NAME} \
  --region ${AWS_REGION} \
  --nodes 2 \
  --node-type t3.large \
  --with-oidc

Configure kubectl Access

Update kubeconfig to point to the new cluster:

aws eks update-kubeconfig \
  --name ${CLUSTER_NAME} \
  --region ${AWS_REGION}

Verify cluster connectivity and node readiness:

kubectl get nodes

If no worker nodes are present, create a managed node group manually:

eksctl create nodegroup \
  --cluster ${CLUSTER_NAME} \
  --region ${AWS_REGION} \
  --name ng-${CLUSTER_NAME} \
  --nodes 2 \
  --node-type t3.large

If nodes fail with CNI plugin not initialized error, install/update the VPC CNI add-on:

Check existing add-ons:

aws eks list-addons \
  --cluster-name ${CLUSTER_NAME} \
  --region ${AWS_REGION}

Install or update the VPC CNI add-on:

aws eks create-addon \
  --cluster-name ${CLUSTER_NAME} \
  --addon-name vpc-cni \
  --region ${AWS_REGION}

Or, if already installed:

aws eks update-addon \
  --cluster-name ${CLUSTER_NAME} \
  --addon-name vpc-cni \
  --region ${AWS_REGION} \
  --resolve-conflicts OVERWRITE

After updating the CNI, delete the failed node group and recreate it:

eksctl delete nodegroup --cluster ${CLUSTER_NAME} --name ng-${CLUSTER_NAME} --region ${AWS_REGION}

eksctl create nodegroup \
  --cluster ${CLUSTER_NAME} \
  --region ${AWS_REGION} \
  --name ng-${CLUSTER_NAME} \
  --nodes 2 \
  --node-type t3.large

Install Essential EKS Add-ons

Ensure the core EKS add-ons required for networking and DNS are installed.

List existing add-ons:

aws eks list-addons \
  --cluster-name ${CLUSTER_NAME} \
  --region ${AWS_REGION}

Install CoreDNS:

aws eks create-addon \
  --cluster-name ${CLUSTER_NAME} \
  --addon-name coredns \
  --region ${AWS_REGION}

Install kube-proxy:

aws eks create-addon \
  --cluster-name ${CLUSTER_NAME} \
  --addon-name kube-proxy \
  --region ${AWS_REGION}

Verify System Components

Confirm all system pods are running:

kubectl get pods -n kube-system

You should see aws-node, coredns, and kube-proxy pods running.