eks-tutorials

Step by step tutorial for who have no experience to Amazon EKS. After finished the tutorial, you should be able to run general workload with Amazon EKS. Hope you enjoy the journey.

Disclaimer

Please note this tutorial is for demonstration purpose only, please DO NOT blindly apply it to your production environments.

Prerequisites

eksctl - The official CLI for Amazon EKS
kubectl - The Kubernetes command-line tool
helm - The Kubernetes Package Manage
Amazon EKS 1.23 or higher - To support autoscaling/v2, learn more at KEP-2702.

`eksctl` version requirements

Click here 🔍

To get support for Amazon EKS 1.31
- support have been added after eksctl-0.191.0 released.
To get support for Amazon EKS 1.30
- support have been added after eksctl-0.179.0 released.
To get support for Amazon EKS 1.29
- support have been added after eksctl-0.169.0 released.
To get support for Amazon EKS 1.28
- support have been added after eksctl-0.160.0 released.
To get support for Amazon EKS 1.27
- support have been added after eksctl-0.143.0 released.
To get support for Amazon EKS 1.26
- support have been added after eksctl-0.138.0 released.
To get support for Amazon EKS 1.25
- support have been added after eksctl-0.132.0 released.
To get support for Amazon EKS 1.24
- support have been added after eksctl-0.120.0 released.
To get support for Amazon EKS 1.23
- support have been added after eksctl-0.109.0 released.
To get support for Amazon EKS 1.22
- support have been added after eksctl-0.92.0 released.
- support have been removed after eksctl-0.151.0 released.

Assumptions

Your AWS Profile have proper permission configured.
All the tools required were setup properly
All the resources are under us-east-1
The cluster name would be eks-demo

Guideline of the tutorial

Goal 1: Create EKS Cluster with eksctl
Goal 2: Deploy nginx with Application Load Balancer (ALB)
Goal 3: Find out why Application Load Balancer (ALB) not working?
Goal 4: Find out why Horizontal Pod Autoscaling (HPA) not working?
Goal 5: HPA is working. Now I want to set Nginx replicas with kubectl scale ... but failed. Why?
Goal 6: Remove HPA and try to scale to 20 manually
Goal 7: Try to turn ALB entry from HTTP to HTTPS
Goal 8: How to switch to Network Load Balancer (NLB)?
Goal 9: Cleanup

Goal 1: Create EKS Cluster with `eksctl`

Make sure you have latest eksctl installed and you should be able to create EKS cluster with minimal setup as follow.

% eksctl create cluster -f ./cluster-config/cluster-minimal.yaml

Click here to show sample deployment output 🔍

2024-XX-XX XX:XX:XX [ℹ]  eksctl version 0.194.0
2024-XX-XX XX:XX:XX [ℹ]  using region us-east-1
2024-XX-XX XX:XX:XX [ℹ]  subnets for us-east-1a - public:192.168.0.0/19 private:192.168.64.0/19
2024-XX-XX XX:XX:XX [ℹ]  subnets for us-east-1b - public:192.168.32.0/19 private:192.168.96.0/19
2024-XX-XX XX:XX:XX [ℹ]  nodegroup "mng-1" will use "" [AmazonLinux2023/1.31]
2024-XX-XX XX:XX:XX [ℹ]  using Kubernetes version 1.31
2024-XX-XX XX:XX:XX [ℹ]  creating EKS cluster "eks-demo" in "us-east-1" region with Fargate profile and managed nodes
2024-XX-XX XX:XX:XX [ℹ]  1 nodegroup (mng-1) was included (based on the include/exclude rules)
2024-XX-XX XX:XX:XX [ℹ]  will create a CloudFormation stack for cluster itself and 0 nodegroup stack(s)
2024-XX-XX XX:XX:XX [ℹ]  will create a CloudFormation stack for cluster itself and 1 managed nodegroup stack(s)
2024-XX-XX XX:XX:XX [ℹ]  if you encounter any issues, check CloudFormation console or try 'eksctl utils describe-stacks --region=us-east-1 --cluster=eks-demo'
2024-XX-XX XX:XX:XX [ℹ]  Kubernetes API endpoint access will use provided values {publicAccess=true, privateAccess=true} for cluster "eks-demo" in "us-east-1"
2024-XX-XX XX:XX:XX [ℹ]  configuring CloudWatch logging for cluster "eks-demo" in "us-east-1" (enabled types: api, audit, authenticator, controllerManager, scheduler & no types disabled)
2024-XX-XX XX:XX:XX [ℹ]
2 sequential tasks: { create cluster control plane "eks-demo",
    2 sequential sub-tasks: {
        4 sequential sub-tasks: {
            1 task: { create addons },
            wait for control plane to become ready,
            update CloudWatch log retention,
            create fargate profiles,
        },
        create managed nodegroup "mng-1",
    }
}
2024-XX-XX XX:XX:XX [ℹ]  building cluster stack "eksctl-eks-demo-cluster"
2024-XX-XX XX:XX:XX [ℹ]  deploying stack "eksctl-eks-demo-cluster"
2024-XX-XX XX:XX:XX [ℹ]  waiting for CloudFormation stack "eksctl-eks-demo-cluster"
2024-XX-XX XX:XX:XX [ℹ]  creating addon
2024-XX-XX XX:XX:XX [ℹ]  successfully created addon
2024-XX-XX XX:XX:XX [ℹ]  creating addon
2024-XX-XX XX:XX:XX [ℹ]  successfully created addon
2024-XX-XX XX:XX:XX [ℹ]  "addonsConfig.autoApplyPodIdentityAssociations" is set to true; will lookup recommended pod identity configuration for "vpc-cni" addon
2024-XX-XX XX:XX:XX [ℹ]  deploying stack "eksctl-eks-demo-addon-vpc-cni-podidentityrole-aws-node"
2024-XX-XX XX:XX:XX [ℹ]  waiting for CloudFormation stack "eksctl-eks-demo-addon-vpc-cni-podidentityrole-aws-node"
2024-XX-XX XX:XX:XX [ℹ]  creating addon
2024-XX-XX XX:XX:XX [ℹ]  successfully created addon
2024-XX-XX XX:XX:XX [ℹ]  creating addon
2024-XX-XX XX:XX:XX [ℹ]  successfully created addon
2024-XX-XX XX:XX:XX [ℹ]  set log retention to 90 days for CloudWatch logging
2024-XX-XX XX:XX:XX [ℹ]  creating Fargate profile "karpenter" on EKS cluster "eks-demo"
2024-XX-XX XX:XX:XX [ℹ]  created Fargate profile "karpenter" on EKS cluster "eks-demo"
2024-XX-XX XX:XX:XX [ℹ]  building managed nodegroup stack "eksctl-eks-demo-nodegroup-mng-1"
2024-XX-XX XX:XX:XX [ℹ]  deploying stack "eksctl-eks-demo-nodegroup-mng-1"
2024-XX-XX XX:XX:XX [ℹ]  waiting for CloudFormation stack "eksctl-eks-demo-nodegroup-mng-1"
2024-XX-XX XX:XX:XX [ℹ]  waiting for the control plane to become ready
2024-XX-XX XX:XX:XX [✔]  saved kubeconfig as "/Users/demoUser/.kube/config"
2024-XX-XX XX:XX:XX [ℹ]  no tasks
2024-XX-XX XX:XX:XX [✔]  all EKS cluster resources for "eks-demo" have been created
2024-XX-XX XX:XX:XX [✔]  created 0 nodegroup(s) in cluster "eks-demo"
2024-XX-XX XX:XX:XX [ℹ]  nodegroup "mng-1" has 2 node(s)
2024-XX-XX XX:XX:XX [ℹ]  node "ip-192-168-121-236.ec2.internal" is ready
2024-XX-XX XX:XX:XX [ℹ]  node "ip-192-168-90-226.ec2.internal" is ready
2024-XX-XX XX:XX:XX [ℹ]  waiting for at least 2 node(s) to become ready in "mng-1"
2024-XX-XX XX:XX:XX [ℹ]  nodegroup "mng-1" has 2 node(s)
2024-XX-XX XX:XX:XX [ℹ]  node "ip-192-168-121-236.ec2.internal" is ready
2024-XX-XX XX:XX:XX [ℹ]  node "ip-192-168-90-226.ec2.internal" is ready
2024-XX-XX XX:XX:XX [✔]  created 1 managed nodegroup(s) in cluster "eks-demo"
2024-XX-XX XX:XX:XX [ℹ]  kubectl command should work with "/Users/demoUser/.kube/config", try 'kubectl get nodes'
2024-XX-XX XX:XX:XX [✔]  EKS cluster "eks-demo" in "us-east-1" region is ready

Verify the EKS nodes are running.

% kubectl get nodes
NAME                              STATUS   ROLES    AGE     VERSION
ip-192-168-121-236.ec2.internal   Ready    <none>   5m47s   v1.30.2-eks-1552ad0
ip-192-168-90-226.ec2.internal    Ready    <none>   5m49s   v1.30.2-eks-1552ad0

Goal 2: Deploy nginx with Application Load Balancer (ALB)

At this stage, you would need to have kubectl installed. Then you should be able to create Deployment, HPA, Service and Ingress resources.

% kubectl apply -f ./examples/simple/
deployment.apps/nginx-deployment created
horizontalpodautoscaler.autoscaling/nginx-hpa created
ingress.networking.k8s.io/nginx-ingress created
service/nginx-service created

Make sure everything run as expected

% kubectl get pods,deployments,hpa,service,ingress
NAME                                    READY   STATUS    RESTARTS   AGE
pod/nginx-deployment-598bb489bf-c55jl   1/1     Running   0          35s
pod/nginx-deployment-598bb489bf-x86dd   1/1     Running   0          35s

NAME                                READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/nginx-deployment    2/2     2            2           35s

NAME                                            REFERENCE                     TARGETS         MINPODS   MAXPODS   REPLICAS   AGE
horizontalpodautoscaler.autoscaling/nginx-hpa   Deployment/nginx-deployment   <unknown>/80%   2         10        2          34s

NAME                    TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
service/kubernetes      ClusterIP   10.100.0.1      <none>        443/TCP        16m
service/nginx-service   NodePort    10.100.94.158   <none>        80:31251/TCP   34s

NAME                                      CLASS   HOSTS                ADDRESS   PORTS   AGE
ingress.networking.k8s.io/nginx-ingress   alb     entry1.example.com             80      35s

Goal 3: Find out why Application Load Balancer (ALB) not working?

% kubectl get ingress nginx-ingress
NAME            CLASS   HOSTS                ADDRESS   PORTS   AGE
nginx-ingress   alb     entry1.example.com             80      50s # <-------- no address shown, why?

After fixing the issue, you should be able to see command output as follow,

% kubectl get ingress nginx-ingress
NAME            CLASS   HOSTS                ADDRESS                                    PORTS   AGE
nginx-ingress   alb     entry1.example.com   k8s-default-XXX.REGION.elb.amazonaws.com   80      60s

Once the Load Balancer is created, you should be able to visit the application via the endpoint of load balancer with default HTTP protocol.

Goal 4: Find out why Horizontal Pod Autoscaling (HPA) not working?

% kubectl get hpa nginx-hpa
NAME        REFERENCE                     TARGETS         MINPODS   MAXPODS   REPLICAS   AGE
nginx-hpa   Deployment/nginx-deployment   <unknown>/80%   2         10        2          68s

Did you aware that HPA is not working... why? 🤔

After you fixed the HPA issue, it should shown as follow

% kubectl get hpa nginx-hpa
NAME        REFERENCE                     TARGETS   MINPODS   MAXPODS   REPLICAS   AGE
nginx-hpa   Deployment/nginx-deployment   2%/80%    2         10        2          2m7s

Goal 5: HPA is working. Now I want to set Nginx replicas with `kubectl scale ...` but failed. Why?

% kubectl scale --replicas 12 deployment nginx-deployment
deployment.apps/nginx-deployment scaled

Why the Pod count not able to reach desired pod count 12 but quickly scale down back to 2... why is that ?

Goal 6: Remove HPA and try to scale to `40` manually

% kubectl delete hpa nginx-hpa
horizontalpodautoscaler.autoscaling "nginx-hpa" deleted

% kubectl scale --replicas 40 deployment nginx-deployment
deployment.apps/nginx-deployment scaled

% kubectl get deployment
NAME               READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   25/40   40           25          11m <-------- stock at "25/40" ...why?

% kubectl get pods
NAME                                READY   STATUS    RESTARTS   AGE
nginx-deployment-598bb489bf-c55jl   1/1     Running   0          6m15s
nginx-deployment-598bb489bf-6tqb8   1/1     Running   0          13m
nginx-deployment-598bb489bf-c55jl   1/1     Running   0          6m15s
nginx-deployment-598bb489bf-d7pcg   0/1     Pending   0          6m15s # <-------- Pending
nginx-deployment-598bb489bf-fw52n   0/1     Pending   0          6m15s # <-------- Pending
nginx-deployment-598bb489bf-x86dd   0/1     Pending   0          6m15s # <-------- Pending
... (omitted)

Goal 7: Try to turn ALB entry from HTTP to HTTPS

Service with HTTP is clearly unsafe, how to made it safe with HTTPS?

Goal 8: How to switch to Network Load Balancer (NLB)?

If you solve can provision ALB then you should be able to create NLB as well. But how...? 🤔

Goal 9: Cleanup

Terminate all resources that we created earlier.

% kubectl delete -f ./examples/simple/ --ignore-not-found

Terminate the EKS cluster

⚠️ WARNING: all resources will be removed permanently, unrecoverable

% eksctl delete cluster -f ./cluster-config/cluster-minimal.yaml

OPTIONAL Cleanup IAM User/Role/Policy and Identity Provider (IdP) with care.

Bonus

There's another repository with common used addons installation scripts:

https://github.com/guessi/eks-integrations-quick-start

Name		Name	Last commit message	Last commit date
Latest commit History 144 Commits
cluster-config		cluster-config
examples		examples
README.md		README.md

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

eks-tutorials

Disclaimer

Prerequisites

`eksctl` version requirements

Assumptions

Guideline of the tutorial

Goal 1: Create EKS Cluster with `eksctl`

Goal 2: Deploy nginx with Application Load Balancer (ALB)

Goal 3: Find out why Application Load Balancer (ALB) not working?

Goal 4: Find out why Horizontal Pod Autoscaling (HPA) not working?

Goal 5: HPA is working. Now I want to set Nginx replicas with `kubectl scale ...` but failed. Why?

Goal 6: Remove HPA and try to scale to `40` manually

Goal 7: Try to turn ALB entry from HTTP to HTTPS

Goal 8: How to switch to Network Load Balancer (NLB)?

Goal 9: Cleanup

Bonus

About

guessi/eks-tutorials

Folders and files

Latest commit

History

Repository files navigation

eks-tutorials

Disclaimer

Prerequisites

eksctl version requirements

Assumptions

Guideline of the tutorial

Goal 1: Create EKS Cluster with eksctl

Goal 2: Deploy nginx with Application Load Balancer (ALB)

Goal 3: Find out why Application Load Balancer (ALB) not working?

Goal 4: Find out why Horizontal Pod Autoscaling (HPA) not working?

Goal 5: HPA is working. Now I want to set Nginx replicas with kubectl scale ... but failed. Why?

Goal 6: Remove HPA and try to scale to 40 manually

Goal 7: Try to turn ALB entry from HTTP to HTTPS

Goal 8: How to switch to Network Load Balancer (NLB)?

Goal 9: Cleanup

Bonus

About

Topics

Resources

Stars

Watchers

Forks

`eksctl` version requirements

Goal 1: Create EKS Cluster with `eksctl`

Goal 5: HPA is working. Now I want to set Nginx replicas with `kubectl scale ...` but failed. Why?

Goal 6: Remove HPA and try to scale to `40` manually