Skip to content


Extend Amazon EKS to AWS Local Zones

Amazon EKS data plane can be extended to be deployed in AWS Local Zones. In this deployment mode the Amazon EKS Kubernetes control plane is deployed in Region but the data plane is extended to the AWS Local Zones. With this configuration, you can deploy applications closer to end-users for low-latency performance while maintaining a centralized management plane for your EKS cluster.

To get started, you need to create an Amazon EKS control plane in an AWS Region. You can then add self-managed nodes to the Amazon EKS cluster that are located in the Local Zones. These nodes can be used to run Kubernetes pods that are designed to serve requests from users in the Local Zone.

To ensure the pods are highly available, you can use Kubernetes' node affinity and anti-affinity features to ensure that pods are scheduled on nodes in the Local Zone and back up pods are scheduled on nodes in the region.

By extending Amazon EKS data plane to AWS Local Zones, you can take advantage of the low-latency benefits of running your application in proximity to end-users while maintaining a centralized management plane for your EKS control plane.

When you are looking to understand how to design workloads that are stretched across AWS Region and Local Zones, this project presents a sample architecture. The project also shares complementary AWS CloudFormation that you can consider to improve operational and developer efficiency.

⚠️ Disclaimer: The sample code; software libraries; command line tools; proofs of concept; templates; or other related technology (including any of the foregoing that are provided by our personnel) is provided to you as AWS Content under the AWS Customer Agreement, or the relevant written agreement between you and AWS (whichever applies). You should not use this AWS Content in your production accounts, or on production or other critical data. You are responsible for testing, securing, and optimizing the AWS Content, such as sample code, as appropriate for production grade use based on your specific quality control practices and standards. Deploying AWS Content may incur AWS charges for creating or using AWS chargeable resources, such as running Amazon EC2 instances or using Amazon S3 storage.

Solution Overview

This solution enables you to simplify and centralize the management of your infrastructure and applications on AWS Region and on AWS Local Zones. You can extend the AWS cloud operations experience across hybrid and Local Zones for secure and seamless management, compliance, and observability. AWS Hybrid Cloud Solutions enable you to deliver a consistent AWS experience wherever you need it—from the cloud, to the edge.

To create an EKS cluster with a mix of managed and self-managed nodes, you can use CloudFormation samples to define and deploy the necessary infrastructure. Here's a high-level overview of the solution:

  • Define the VPC and Subnets
  • Deploy the EKS Cluster
  • Add Self-Managed Nodes
  • Deploy sample applications in the Region and Local Zones

The following diagram shows the high-level architecture, for running a sample website on Amazon EKS in the Local Zone and the Region.

You will be deploying to sample applications one in the Local Zones and another backup one in the Region

When you are connecting to the Local Zone, the request is served by the ALB in the Local Zone, and the game (sample application) is hosted by Kubernetes pods, running on the self-managed Amazon EC2 nodes.

For the backup site in the Region, there is an ALB and Kubernetes pods running on a managed node group. The backup game is used for High Availability that makes it easy for IT administrators to set up, operate, and scale in the cloud.


Amazon EKS Cluster Deployment using Local Zones

For the application deployment, we use the combination of CloudFormation YAML files. We use CloudFormation to create AWS resources such as Amazon Virtual Private Cloud (Amazon VPC), Amazon EKS, Amazon EC2, etc. For the application in Kubernetes, we use the YAML manifest files and 2048 game available in this Solution.

Note Important note for developers: All commands mentioned in the instructions were tested using Cloud9 using the Amazon Linux 2 as a platform. If you encounter any issues while working with a different operating system, please verify the applicability of the command to your specific operating system. Some commands may have variations or different syntax depending on the OS in use. Ensure compatibility and make any necessary adjustments before executing the commands on your system.


  • An AWS account, IAM user hybrid-eks-user and a role hybrid-eks-user-role with Administrator permissions.

Note It is important to create the Amazon EKS cluster with a dedicated IAM role and regularly audit who can assume this role.

  • A shell environment. An IDE (Integrated Development Environment) environment such as Visual Studio Code or AWS Cloud9 is recommended.

  • Installation of the latest version AWS Command Line Interface (AWS CLI) (v2 recommended), kubectl, eksctl, and EC2 Key Pair

#Download and extract the latest release of awscli with the following
curl "" -o ""
sudo ./aws/install

# Download and extract the latest release of eksctl with the following
curl --silent --location "$(uname -s)_amd64.tar.gz" | tar xz -C /tmp
sudo mv /tmp/eksctl /usr/local/bin
eksctl version

# Install kubectl for the self-managed nodes at the Local Zones side.
curl -o kubectl
openssl sha1 -sha256 kubectl
chmod +x ./kubectl
mkdir -p $HOME/bin && cp ./kubectl $HOME/bin/kubectl && export PATH=$HOME/bin:$PATH
echo 'export PATH=$PATH:$HOME/bin' >~/.bash_profile
kubectl version --short --client

# Install jq
sudo yum install -y jq

# Install helm
curl -fsSL -o
chmod 700

# Create an EC2 key pair [optional], if you decided to use existing one please make sure to update the CloudFormation parameter KeyPairName-
aws ec2 create-key-pair --key-name eks-keypair --region us-west-2 --query 'KeyMaterial' --output text > MyKeyPair.pem
  • To assume role, using AWS Identity and Access Management (AWS IAM) Role, the following commands help to assume role as you need to configure IAM user hybrid-eks-user policy, update role hybrid-eks-user-role trust policy then assume role.
account_id=$(aws sts get-caller-identity --query Account --output text)
echo $account_id
#create user policy enabling the user to assumer role 
cat >user-policy.json <<EOF
    "Version": "2012-10-17",
    "Statement": [
        "Effect": "Allow",
        "Action": [
        "Resource": "*"

#create policy
aws iam create-policy --policy-name hybrid-eks-policy --policy-document file://user-policy.json

#attach policy 
aws iam attach-user-policy --user-name hybrid-eks-user --policy-arn "arn:aws:iam::$account_id:policy/hybrid-eks-policy"

cat >eks-role-trust-policy.json <<EOF
    "Version": "2012-10-17",
    "Statement": {
      "Effect": "Allow",
      "Principal": {
        "AWS": "$account_id"
      "Action": "sts:AssumeRole"

#update hybrid-eks-user-role role assume role policy 
aws iam update-assume-role-policy --role-name hybrid-eks-user-role --policy-document file://eks-role-trust-policy.json

#reconfigure aws cli token for the user `hybrid-eks-user`
aws configure
  • Assume role
# assume role
credentials=$(aws sts assume-role --duration-seconds 3600 --role-arn $role_arn --role-session-name eks)
# Set the AWS credentials as environment variables
export AWS_ACCESS_KEY_ID=$(echo "$credentials" | jq -r '.Credentials.AccessKeyId')
export AWS_SECRET_ACCESS_KEY=$(echo "$credentials" | jq -r '.Credentials.SecretAccessKey')
export AWS_SESSION_TOKEN=$(echo "$credentials" | jq -r '.Credentials.SessionToken')
export AWS_EXPIRATION=$(echo "$credentials" | jq -r '.Credentials.Expiration')
  • Opt-in for us-west-2-lax-1a Local Zone that this sample will create and run some workloads in. Please refer to the docs here to enable LAX Local Zone, Region us-west-2.


Step 1: Create the topology along EKS cluster in the Region

  • Initiate the CloudFormation stack using eks-cluster YAML file:
 aws cloudformation create-stack \
  --stack-name eks-cp-cfn-stack \
  --template-body file://eks-cluster.yaml \
  --parameters ParameterKey=EKSClusterName,ParameterValue=eks-cluster ParameterKey=NumWorkerNodes,ParameterValue=2  ParameterKey=KeyPairName,ParameterValue=eks-keypair ParameterKey=Region,ParameterValue=us-west-2 ParameterKey=LocalZones,ParameterValue=us-west-2-lax-1a
 aws cloudformation wait stack-create-complete --stack-name eks-cp-cfn-stack

When you observe the StackId prompt appearing in your terminal, it indicates that the execution of your CloudFormation has been initiated successfully. This prompt serves as a confirmation that your CloudFormation stack has started to execute.

  • Wait for 10-15 mins for the CloudFormation to create all resources "CREATE_COMPLETE" status for the stack.

Step 2: Deploy and setup the self-managed node

  • Prepare the environment for Kubernetes and self-managed node
#Ensure that kubectl is configured for the Region and the newly created cluster.
aws eks update-kubeconfig \
            --region us-west-2 \
            --name eks-cluster 

#Fetch required resources generated through the previous step
cluster_cp_security_group=$(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`EKSClusterSG`].OutputValue' --output text)

key_name=$(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`KeyPair`].OutputValue' --output text)

subnets=$(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`LZPrivateSubnetId`].OutputValue' --output text)

vpc=$(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`VpcId`].OutputValue' --output text)
  • Create the self-managed nodes
aws cloudformation create-stack \
  --stack-name eks-lz-nodes-cfn-stack \
  --template-body file://eks-selfmanaged-node.yaml \
  --parameters ParameterKey=ClusterName,ParameterValue=eks-cluster ParameterKey=ClusterControlPlaneSecurityGroup,ParameterValue=$cluster_cp_security_group  ParameterKey=KeyName,ParameterValue=$key_name ParameterKey=Subnets,ParameterValue=$subnets ParameterKey=VpcId,ParameterValue=$vpc ParameterKey=NodeGroupName,ParameterValue=eks-selfmanaged-groupnode ParameterKey=NodeImageId,ParameterValue=ami-0b149b4c68ab69dce
aws cloudformation wait stack-create-complete --stack-name eks-lz-nodes-cfn-stack

Note When you observe the StackId prompt appearing in your terminal, it indicates that the execution of your CloudFormation has been initiated successfully. This prompt serves as a confirmation that your CloudFormation stack has started to execute.

  • Wait for 5 mins for the CloudFormation to create all resources "CREATE_COMPLETE" status for the stack.

  • Enable self-managed nodes to join your cluster, as the following:

  1. Map the newly added rolearn to the EKS cluster and apply the configuration to Amazon IAM Authenticator configuration map using the following commands. This process may take a few minutes to complete.
 lz_node_role=$(aws cloudformation describe-stacks --stack-name "eks-lz-nodes-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`NodeInstanceRole`].OutputValue' --output text)

 eksctl create iamidentitymapping --cluster eks-cluster --region=us-west-2 --arn $lz_node_role --group system:bootstrappers --group system:nodes --username "system:node:{{EC2PrivateDNSName}}"
  1. Verify the Amazon IAM Authenticator configuration map details
  • Get authenticator configuration map details
kubectl get -n kube-system configmap/aws-auth -o yaml  > aws-auth-cm.yaml 
  • Open the aws-auth-cm.yaml file in the left panel of your Cloud9 IDE.

  • Make sure the rolearn of the newly added group with the lz-nodes role is there as the following:


  1. Watch the status of your nodes and wait for them to reach the Ready status.
 kubectl get nodes --watch

To stop monitoring the status of nodes in a Kubernetes cluster using the "watch" command, you can press "Ctrl + C" when you observe that new node is in the "Ready" state. This will terminate the command and exit the watch mode.


Warning If you receive any authorization or resource type errors, see Unauthorized or access denied (kubectl) and further references in the troubleshooting topic.

Step 3: Installing the AWS Load Balancer Controller add-on

  1. Creating an AWS Identity and Access Management (IAM) OpenID Connect (OIDC) provider for your cluster.

Amazon EKS supports using OpenID Connect (OIDC) identity providers as a method to authenticate users to your cluster, further details

  • Retrieve your cluster's OIDC provider ID and store it in a variable.
 oidc_id=$(aws eks describe-cluster --name eks-cluster --query "cluster.identity.oidc.issuer" --output text | cut -d '/' -f 5)
 account_id=$(aws sts get-caller-identity --query Account --output text)
  • Create an IAM OIDC identity provider for your cluster with the following command.
 eksctl utils associate-iam-oidc-provider --cluster eks-cluster --approve
  1. Deploy the AWS Load Balancer Controller to an Amazon EKS cluster
  • Create an IAM policy using the policy downloaded in the previous step.
 curl -O
 aws iam create-policy \
    --policy-name AWSLoadBalancerControllerIAMPolicy \
    --policy-document file://iam_policy.json
  • Create an IAM role. Create a Kubernetes service account named aws-load-balancer-controller in the kube-system namespace for the AWS Load Balancer Controller and annotate the Kubernetes service account with the name of the IAM role.
eksctl create iamserviceaccount \
 --cluster=eks-cluster \
 --namespace=kube-system \
 --name=aws-load-balancer-controller \
 --role-name AmazonEKSLoadBalancerControllerRole \
 --attach-policy-arn=arn:aws:iam::$account_id:policy/AWSLoadBalancerControllerIAMPolicy \
  • Install the AWS Load Balancer Controller, as the following:
#Add the eks-charts repository.
helm repo add eks
#Update your local repo to make sure that you have the most recent charts.
helm repo update eks
#get vpcId
vpc_id=$(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`VpcId`].OutputValue' --output text)
#Install the AWS Load Balancer Controller.
helm install aws-load-balancer-controller eks/aws-load-balancer-controller -n kube-system --set clusterName=eks-cluster --set serviceAccount.create=false --set --set vpcId=$vpc_id --set region=us-west-2
  • Verify that the controller is installed.
 kubectl get deployment -n kube-system aws-load-balancer-controller
 kubectl get pod -n kube-system

The example output is as follows.

NAME                           READY   UP-TO-DATE   AVAILABLE   AGE
aws-load-balancer-controller   1/1     1            1          84s


Step 4: Deploy two sample application to the Region and to the Local Zones

  1. Deploy the game 2048 as a sample application to verify that the AWS Load Balancer Controller creates an AWS ALB as a result of the ingress object or use the sample games configured along the nodeAffinity in this solution.

  2. Get the public subnet in the Local Zones

 #deploying to pods in a cluster in the Region (public)
 echo $(aws cloudformation describe-stacks --stack-name "eks-cp-cfn-stack" --region "us-west-2" --query 'Stacks[0].Outputs[?OutputKey==`LZPublicSubnetId`].OutputValue' --output text)
  1. Open 2048_lz file and update tag "" with subnet captured from previous step, save file, then deploy the applications using the following command
#Deploying to pods in a cluster in the Region (public)
 kubectl apply -f 2048_backup.yaml
#Deploying to pods in a cluster in the Local Zones (public)
 kubectl apply -f 2048_lz.yaml

Note Open 2048_lz.yaml file and search for nodeAffinity, as you can see by setting up the affinity rules in the pod configuration using the nodeAffinity field, we specified that the pod should be scheduled only on nodes with the zone.

  1. After a few minutes, verify that the ingress resource was created with the following command.
 kubectl get ingress/ingress-2048-backup -n game-2048-backup
 kubectl get ingress/ingress-2048-lz -n game-2048-lz

The example output is as follows.

NAME                  CLASS   HOSTS   ADDRESS                                                                  PORTS   AGE
ingress-2048-backup   alb     *   80      7m
ingress-2048-lz   alb     *   80      2m
  1. To verify successful installation, open a browser and navigate to the ADDRESS URL from the previous commands output to see the sample application. If you don't see anything, refresh your browser and try again or troubleshoot .



  • Kubernetes assigns the service its own IP address that is accessible only from within the cluster. To access the service from outside of your cluster, deploy the AWS Load Balancer Controller to load balance application or network traffic to the service.

  • Enabling IAM user and role access to your cluster

  • Installing the AWS Load Balancer Controller add-on further details


If you encounter a situation where you are unable to obtain an address for a test games, it may be helpful to troubleshoot the AWS Load Balancer Controller. One possible solution is to delete the controller using the full pod name. To do this, you can run the following command:

kubectl get pods -o wide -n kube-system

Replace with the full name of the AWS Load Balancer Controller pod.

kubectl delete pods <aws-loadbalancer-controller-pod> -n kube-system

If this didn't work as well, you can use the kubectl describe command with the appropriate resource name, such as service or ingress, followed by the name of the load balancer controller. This command provides detailed information about the load balancer controller's configuration, status, and any associated events that may indicate issues or errors. Additionally, you can retrieve events to check the status.

kubectl describe deploy aws-load-balancer-controller -n kube-system | grep -i "Service Account"
kubectl get events -w

To stop monitoring the status of events in a Kubernetes cluster using the "watch" command, you can press "Ctrl + C". This will terminate the command and exit the watch mode.

Optional Step 5: Configure Route 53 failover for high availability

Assuming you have a public domain defined under Amazon Route 53 hosted zones, we have a complementary CloudFormation stack that can help you create failover alias records values, configure Amazon Route 53 health checks, and set up DNS failover.

To configure the CloudFormation stack, we require some input details. Firstly, please provide the Hosted Zone ID for the domain name. This ID represents the domain for which you have administrative control and can manage DNS records. Once we have the Hosted Zone ID, we can proceed with the configuration.

In addition, we need you to input the LZ DNS name. This should correspond to the ingress resource address we created in the previous step, specifically for get ingress/ingress-2048-lz. The LZ DNS name will be used for the primary failover record set. Please ensure to include a period (.) at the end of the LZ DNS name to make it fully qualified.

Similarly, we need the backup DNS name. This should correspond to the ingress resource address we created in the previous step for ingress/ingress-2048-backup. The backup DNS name will serve as the backup failover record set. Remember to include a period (.) at the end of the backup DNS name as well.

By providing these inputs myhostZoneId,,PDNS and SDNS accurately, you can proceed with creating the CloudFormation stack to set up the failover alias records, configure health checks, and enable DNS failover for your domain.


aws cloudformation create-stack \
  --stack-name eks-route53-cfn-stack \
  --template-body file://eks-route53.yaml \
  --parameters ParameterKey=HostedZoneId,ParameterValue=$HostedZoneId ParameterKey=DomainName,ParameterValue=$DomainName  ParameterKey=PrimaryALBDNS,ParameterValue=$PrimaryALBDNS. ParameterKey=SecondaryALBDNS,ParameterValue=$SecondaryALBDNS.

Verify high availability deployment

aws cloudformation wait stack-create-complete --stack-name eks-route53-cfn-stack
echo $GAME_URL
while true; do response=$(curl -s -o /dev/null -w "%{http_code}" -m 2 $GAME_URL); [ "$response" == "200" ] && echo "Webpage is responding OK" || echo "Webpage is not responding"; sleep 1; done

To stop verification when the status is 'Webpage is not responding', you can press "Ctrl + C". This will terminate the command and exit the loop.

Lastly, scale down the Local Zone's nodes so that you verify the back up game.

kubectl scale -n game-2048-lz deployment —replicas 0 deployment-2048-lz
deployment.apps/deployment-2048-lz scaled

#retest the game again!
while true; do response=$(curl -s -o /dev/null -w "%{http_code}" -m 2 $GAME_URL); [ "$response" == "200" ] && echo "Webpage is responding OK" || echo "Webpage is not responding"; sleep 1; done

Clean up

To terminate the resources that we created in this sample, as the following:

  • Detach polices from created roles eks-lz-nodes-cfn-stack-NodeInstanceRole-XXXXXX, eks-cp-cfn-stack-WorkerNodesRole-YYYYYY and eks-cp-cfn-stack-ControlPlaneRole-ZZZZZZ.

  • Run the following:

kubectl delete -f 2048_backup.yaml
kubectl delete -f 2048_lz.yaml
helm delete aws-load-balancer-controller -n kube-system

then delete the CloudFormation stacks as the following:

aws cloudformation delete-stack --stack-name eksctl-eks-cluster-addon-iamserviceaccount-kube-system-aws-load-balancer-controller
aws cloudformation wait stack-delete-complete --stack-name eksctl-eks-cluster-addon-iamserviceaccount-kube-system-aws-load-balancer-controller

aws cloudformation delete-stack --stack-name eks-route53-cfn-stack
aws cloudformation wait stack-delete-complete --stack-name eks-route53-cfn-stack

aws cloudformation delete-stack --stack-name eks-lz-nodes-cfn-stack
aws cloudformation wait stack-delete-complete --stack-name eks-lz-nodes-cfn-stack

aws cloudformation delete-stack --stack-name eks-cp-cfn-stack
aws cloudformation wait stack-delete-complete --stack-name eks-cp-cfn-stack

Then, go to the Cloudformation console and make sure the stacks were deleted. Lastly, delete created user, policy and role assumed.

  • Delete AWSLoadBalancerControllerIAMPolicy policy.


Extending an Amazon EKS cluster to AWS Local Zones offers many benefits for organizations looking to improve the performance, availability, and resiliency of their containerized applications. With this configurations, you can leverage the low-latency access to compute and storage resources in geographically closer locations to your end-users or data sources.

In this solution, we have covered the essential steps to extend an Amazon EKS cluster to Local Zones, this includes setting up Amazon EKS cluster, launching self-managed node in the Local Zone, joining the node in the Local Zone to the EKS cluster and deploying a sample application.

By following these steps, you leveraged an Amazon EKS cluster that is highly available, fault-tolerant, and scalable while maintaining a centralized control plane. With this infrastructure, you can easily deploy and manage containerized applications in multiple locations, improving the user experience and reducing the risk of downtime.


See CONTRIBUTING for more information.


This library is licensed under the MIT-0 License. See the LICENSE file.


No description, website, or topics provided.



Code of conduct

Security policy





No releases published


No packages published