You will need to have installed the AWS CLI installed and configured, with the appropriate credentials to be able to create EKS clusters and create S3 buckets and policies. You will also need to have installed eksctl.
To create a cluster, run the following commands, making any necessary adjustment to the variables:
export AWS_PROFILE= # optional
export AWS_REGION=us-east-2
export K8S_CLUSTER_NAME=stackgres
eksctl --region $AWS_REGION create cluster \
--name $K8S_CLUSTER_NAME \
--node-type m5a.2xlarge --node-volume-size 100 --nodes 3 \
--zones ${AWS_REGION}a,${AWS_REGION}b,${AWS_REGION}c \
--version 1.17
This process takes around 20 minutes. The output should be similar to:
2021-02-28 20:26:56 [ℹ] eksctl version 0.39.0
2021-02-28 20:26:56 [ℹ] using region us-east-2
2021-02-28 20:26:56 [ℹ] subnets for us-east-2a - public:192.168.0.0/19 private:192.168.96.0/19
2021-02-28 20:26:56 [ℹ] subnets for us-east-2b - public:192.168.32.0/19 private:192.168.128.0/19
2021-02-28 20:26:56 [ℹ] subnets for us-east-2c - public:192.168.64.0/19 private:192.168.160.0/19
2021-02-28 20:26:57 [ℹ] nodegroup "ng-092df631" will use "ami-0ba2dda6dd6a9e644" [AmazonLinux2/1.17]
2021-02-28 20:26:58 [ℹ] using Kubernetes version 1.17
2021-02-28 20:26:58 [ℹ] creating EKS cluster "stackgres" in "us-east-2" region with un-managed nodes
2021-02-28 20:26:58 [ℹ] will create 2 separate CloudFormation stacks for cluster itself and the initial nodegroup
2021-02-28 20:26:58 [ℹ] if you encounter any issues, check CloudFormation console or try 'eksctl utils describe-stacks --region=us-east-2 --cluster=stackgres'
2021-02-28 20:26:58 [ℹ] CloudWatch logging will not be enabled for cluster "stackgres" in "us-east-2"
2021-02-28 20:26:58 [ℹ] you can enable it with 'eksctl utils update-cluster-logging --enable-types={SPECIFY-YOUR-LOG-TYPES-HERE (e.g. all)} --region=us-east-2 --cluster=stackgres'
2021-02-28 20:26:58 [ℹ] Kubernetes API endpoint access will use default of {publicAccess=true, privateAccess=false} for cluster "stackgres" in "us-east-2"
2021-02-28 20:26:58 [ℹ] 2 sequential tasks: { create cluster control plane "stackgres", 2 sequential sub-tasks: { wait for control plane to become ready, create nodegroup "ng-092df631" } }
2021-02-28 20:26:58 [ℹ] building cluster stack "eksctl-stackgres-cluster"
2021-02-28 20:26:59 [ℹ] deploying stack "eksctl-stackgres-cluster"
2021-02-28 20:39:53 [ℹ] waiting for CloudFormation stack "eksctl-stackgres-cluster"
2021-02-28 20:39:56 [ℹ] building nodegroup stack "eksctl-stackgres-nodegroup-ng-092df631"
2021-02-28 20:39:56 [ℹ] --nodes-min=3 was set automatically for nodegroup ng-092df631
2021-02-28 20:39:56 [ℹ] --nodes-max=3 was set automatically for nodegroup ng-092df631
2021-02-28 20:39:57 [ℹ] deploying stack "eksctl-stackgres-nodegroup-ng-092df631"
2021-02-28 20:44:24 [ℹ] waiting for CloudFormation stack "eksctl-stackgres-nodegroup-ng-092df631"
2021-02-28 20:44:25 [ℹ] waiting for the control plane availability...
2021-02-28 20:44:25 [✔] saved kubeconfig as "/home/aht/.kube/config"
2021-02-28 20:44:25 [ℹ] no tasks
2021-02-28 20:44:25 [✔] all EKS cluster resources for "stackgres" have been created
2021-02-28 20:44:26 [ℹ] adding identity "arn:aws:iam::292778140943:role/eksctl-stackgres-nodegroup-ng-092-NodeInstanceRole-1ITJQEMJMGD8" to auth ConfigMap
2021-02-28 20:44:26 [ℹ] nodegroup "ng-092df631" has 0 node(s)
2021-02-28 20:44:26 [ℹ] waiting for at least 3 node(s) to become ready in "ng-092df631"
2021-02-28 20:44:52 [ℹ] nodegroup "ng-092df631" has 3 node(s)
2021-02-28 20:44:52 [ℹ] node "ip-192-168-13-0.us-east-2.compute.internal" is ready
2021-02-28 20:44:52 [ℹ] node "ip-192-168-53-249.us-east-2.compute.internal" is ready
2021-02-28 20:44:52 [ℹ] node "ip-192-168-94-213.us-east-2.compute.internal" is ready
2021-02-28 20:44:53 [✔] EKS cluster "stackgres" in "us-east-2" region is ready
Once your EKS cluster is created, you should have your ~/.kube/config populated, being able to run:
kubectl cluster-info
and get an output similar to:
Kubernetes control plane is running at https://6E48B1E2BBDE5960F174FD2D04C1F554.gr7.us-east-2.eks.amazonaws.com
CoreDNS is running at https://6E48B1E2BBDE5960F174FD2D04C1F554.gr7.us-east-2.eks.amazonaws.com/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'.