container.training/slides/k8s/access-eks-cluster.md

Accessing our EKS cluster

• We also have a shared EKS cluster

• With individual IAM users

• Let's connect to this cluster!

---

What we need

• kubectl (obviously!)

• aws CLI (recent-ish version)

  (or aws CLI + aws-iam-authenticator plugin)

• AWS API access key and secret access key

• AWS region

• EKS cluster name

---

Setting up AWS credentials

• There are many ways to do this

• We're going to use environment variables

• You're welcome to use whatever you like (e.g. AWS profiles)

.lab[

• Set the AWS region, API access key, and secret key:

  export AWS_DEFAULT_REGION=`us-east-2`
  export AWS_ACCESS_KEY_ID=`AKI...`
  export AWS_SECRET_ACCESS_KEY=`xyz123...`
  

• Check that the AWS API recognizes us:

  aws sts get-caller-identity
  

]

---

Updating our kubeconfig file

• Now we can use the AWS CLI to:

  • obtain the Kubernetes API address

  • register it in our kubeconfig file

.lab[

• Update our kubeconfig file:

  aws eks update-kubeconfig --name `fancy-clustername-1234`
  

• Run some harmless command:

  kubectl version
  

]

---

Our resources

• We have the following permissions:

  • view in the default namespace

  • edit in the container-training namespace

  • admin in our personal namespace

• Our personal namespace is our IAM user name

  (but with dots replaced with dashes)

• For instance, user ada.lovelace has namespace ada-lovelace

---

Deploying things

• Let's deploy DockerCoins in our personal namespace!

• Expose the Web UI with a LoadBalancer service

???

:EN:- Working with an EKS cluster :FR:- Travailler avec un cluster EKS