Skip to content
This repository has been archived by the owner on Apr 7, 2023. It is now read-only.

Latest commit

 

History

History
171 lines (96 loc) · 4.59 KB

README.md

File metadata and controls

171 lines (96 loc) · 4.59 KB
Raw

Traefik ingress with Amazon EKS

This guide will walk you through building a Amazon EKS cluster with Traefik Ingress with ALB for HTTPS termionation.

  1. Create Amazon EKS cluster with eksctl - https://github.com/pahud/amazon-eks-workshop/blob/master/00-getting-started/create-eks-with-eksctl.md

  2. Update your NodeGroup with this cloudformation template - https://github.com/pahud/amazon-eks-workshop/blob/master/01-nodegroup/customize-nodegroup.md. (This will also create a Target Group and AutoScaling Group registering to this Target Group on port 31742, on which we will spin up our Traefik Ingress Service to listen and expose as NodePort service type.)

  3. kubectl apply all the yaml in this folder:

    $ cd traefik-ingress
$ kubectl apply -f .
service "greeting-service" created
deployment.extensions "greeting" created
service "caddy-service" created
deployment.extensions "caddy" created
service "nginx-service" created
deployment.extensions "nginx" created
serviceaccount "ingress" created
clusterrolebinding.rbac.authorization.k8s.io "ingress" created
ingress.extensions "traefik-ingress" created
service "traefik-ingress-lb-svc" created
daemonset.extensions "traefik-ingress-lb" created

  4. kubectl get svc to list all the services make sure caddy-service, greeting-service and nginx-service are running

  5. kubectl get ing and make your traefik-ingress is running

  6. kubectl get svc -n kube-system and make sure traefik-ingress-lb-svc is running on NodePort 31742 mapping to container port 80

Describe the ingress and make sure the ingress rules exist

Create an ALB manually and make sure to associate it with two security groups:

  1. The EKS nodegroup securigy group - so ALB can access any port for health check on the nodes

  2. HTTP 80 and HTTPS 443 public open to all

And create a HTTP listener with a forwarding rule to the Target Group on Port 31742

You should be able to see all the EC2 instances registered to this target group become status healthy.

Validate

$ curl -s http://<ALB_DNS_NAME>/caddy| grep title | head -n1
<title>Caddy</title>

$ curl -s http://<ALB_DNS_NAME>/nginx | grep title | head -n1
<title>Welcome to nginx!</title>

$ curl -s http://<ALB_DNS_NAME>/?name=pahud
Hello pahud

Traefik Ingerss Controller running on onDemand EC2

In your hybrid Amazon EKS nodegroup, your ondemand instances may have a label asgnode=true and spot instances with a label spotfleet=true. You can get them like this:

In traefik-ingress-lb.yaml we have a nodeSelector as below to restrict the traefik ingress controller will only run as DaemonSet on on demand instances managed by autoscaling group.

nodeSelector:
  asgnode: "true"

Get all the traefik ingress controller pods. Please note they only run on on demand instances.

$ kubectl get po -n kube-system -l name=traefik-ingress-lb -o wide
NAME                       READY     STATUS    RESTARTS   AGE       IP                NODE
traefik-ingress-lb-4djwc   1/1       Running   0          15m       192.168.102.191   ip-192-168-102-191.us-west-2.compute.internal
traefik-ingress-lb-x9lts   1/1       Running   0          15m       192.168.142.221   ip-192-168-142-221.us-west-2.compute.internal

Traefik admin dashboard

Get the 1st Pod of traefik ingress controller

podName=$(kubectl -n kube-system get po -l name=traefik-ingress-lb -o jsonpath='{.items[0].metadata.name}')

And check the value of the podName

$ echo $podName
traefik-ingress-lb-9pc2w

(your podName may have different suffix)

Create a port-forward from localhost:8580 into the pod:8580

$ kubectl -n kube-system port-forward pod/$podName 8580:8580
Forwarding from 127.0.0.1:8580 -> 8580
Forwarding from [::1]:8580 -> 8580
Handling connection for 8580

Open http://localhost:8580 from your local browser and you got the dashboard now!

Clean up

$ cd traefik-ingress
$ kubectl delete -f .
service "greeting-service" deleted
deployment.extensions "greeting" deleted
service "caddy-service" deleted
deployment.extensions "caddy" deleted
service "nginx-service" deleted
deployment.extensions "nginx" deleted
serviceaccount "ingress" deleted
clusterrolebinding.rbac.authorization.k8s.io "ingress" deleted
ingress.extensions "traefik-ingress" deleted
service "traefik-ingress-lb-svc" deleted
daemonset.extensions "traefik-ingress-lb" deleted