EKS-deployment

Description

This document provides a process to generate all the required pre-requisite resources via AWS CLI and the EKS resources via EKSCTL and KUBECTL.

Pre-Requisites

You will need to have an AWS IAM user with admin privileges and CLI access, as well as aws-cli, kubectl, and eksctl installed on either your local machine or an EC2 instance to run the appropriate commands to perform an EKS cluster deployment using the methods to follow. eksctl makes it very easy to create and manage eks clusters, creating the cloudformation stack and all necessary resources for your cluster within a single command.

AWS IAM User

Via aws-cli

aws iam create-user --user-name <user-name>
aws iam create-access-key --user-name <user-name>
NOTE: Save AccessKeyId and SecretAccessKey from the output from aws iam create-access-key
aws iam attach-user-policy --user-name <user-name> --policy-arn arn:aws:iam::aws:policy/AdministratorAccess
NOTE: The above provides full admin access to the user. If you wish to create a user with more restricted access rights, you may create a custom IAM policy and attach it via the ARN, replacing the --policy-arn value above.

Using an amazon linux EC2 instance

Create an EC2 Instance

Via aws-cli

Create key-pair

aws ec2 create-key-pair --key-name my-k8-server-key --query 'KeyMaterial' --output text > my-k8-server-key.pem

Update the key's permissions.

Linux: chmod 400 my-k8-server-key

Windows: Right-click on the key file and go to properties. Navigate to Security > Advanced 'Disable inheritance' if it is enabled. Remove all permission entries, except for 'Full control' access for your user.

Create Security Group

Retrieve the VPC ID that you will need this security group deployed for. To retrieve the VPC by VPC name and region: aws ec2 describe-vpcs --filters Name='tag:Name',Values='<VPC NAME>' --region <REGION> --output=text | find /i "vpc-" NOTE: Replace and appropriately. Record the vpc-id from the output. aws ec2 create-security-group --group-name ssh-sg --description "Allow SSH" --vpc-id <VPC ID> NOTE: Replace the from the previous command. Record the GroupId from the output.

aws ec2 authorize-security-group-ingress --group-id <your_security_group_id> --protocol tcp --port 22 --cidr 0.0.0.0/0 NOTE: You can limit the --cidr field to your own public IPV4 address to limit access.

Get Latest Amazon Linux 2 AMI

aws ec2 describe-images --owners amazon --filters "Name=name,Values=amzn2-ami-hvm-2.0.????????.?-x86_64-gp2" "Name=state,Values=available" "Name=virtualization-type,Values=hvm" --region <REGION> --query "reverse(sort_by(Images, &CreationDate))[0].ImageId" NOTE: Replace appropriately. Record the output to use in the EC2 instance creation for the AMI image to use.

Launch the EC2 Instance

TO-DO: Below command will not work without specifying a subnet from a VPC you wish to launch the instance into. aws ec2 run-instances --image-id <AMI_ID> --count 1 --instance-type t2.micro --key-name my-k8-server-key --security-group-ids <your_security_group_id>' NOTE: Replace <AMI_ID> from the output of the previous step. Replace Replace <your_security_group_id> from the output of the Create Security Group step.

Installing/Upgrading aws-cli

https://docs.aws.amazon.com/cli/latest/userguide/getting-started-install.html
Instructions on installing can be found in the above link (subject to change, so will not document exact steps here).
NOTE: There are times when running aws --version shows the older version. After you perform the update, close and reconnect to the EC2 instance and re-run aws --version.

Installing kubectl

https://docs.aws.amazon.com/eks/latest/userguide/install-kubectl.html
Instructions on installing can be found in the above link (subject to change, so will not document exact steps here).
NOTE: Ensure to select the appropriate kubectl version with kubernetes cluster version you will be managing.

Installing eksctl

https://docs.aws.amazon.com/eks/latest/userguide/eksctl.html
Instructions on installing can be found in the above link (subject to change, so will not document exact steps here).

Using eksctl to Deploy a Cluster

Create EKS Cluster

From either your computer or the EC2 you've setup to deploy the cluster from, you can use eksctl to perform a cluster deployment. Below is an example command:
eksctl create cluster --name dev --region us-east-1 --nodegroup-name standard-workers --node-type t3.medium --nodes-min 1 --nodes-max 4 --managed
NOTE: Replace the fields accordingly with your own values.
This will create a cloud-formation stack with a name with format eksctl-<name>-cluster
This cloud formation stack will take approximately 15 minutes to deploy all necessary resources for the eks cluster.
NOTE: Sometimes the deployment will fail due to a lack of resources for the region you're deploying the cluster in. If this occurs, you will need to specify AZs that were not included in the error message to the eksctl command above.

Configure kubectl With Your Cluster

Once the cloud formation stacks have Run the following to connect and configure kubectl with your eks cluster:
aws eks update-kubeconfig --name <CLUSTER NAME> --region <REGION>

Setup The Service

kubectl apply -f ./<service config yaml file>
Confirm the service creation by running: kubectl get service

Setup The Deployment

kubectl apply -f ./<deployment config yaml file>
Confirm the deployment has been completed:
kubectl get deployment
You can also check the status of the pods, replica set, and nodes via:
kubectl get pod
kubectl get rs
kubectl get node

Test The App

If deploying your app with a load balancer as the service, when you run kubectl get service, the load balancer public address will be shown. You can curl, or alternatively utilize your browser to confirm the app is being served.