Everything in these Kubernetes docs works on Amazon EKS. The EKS guide lives with the rest of the AWS docs:
It covers creating an Auto Mode cluster in the same Stack, granting
access, and what EKS adds when you pass the cluster as the cluster
of a Kubernetes.* resource:
- No kubeconfig. Alchemy authenticates with your AWS credentials.
- A registry per workload.
mainandcontextworkloads are built into ECR, so noregistrysetting is needed. - AWS bindings. Bindings such as
AWS.DynamoDB.PutItem(table)grant IAM permissions to the workload through EKS Pod Identity. - Public load balancers.
LoadBalancerServices get an internet-facing Network Load Balancer.
To get started with AWS, see AWS setup.
Use an EKS cluster through kubeconfig
A context written by aws eks update-kubeconfig works with
Kubernetes.KubeConfig like any
other cluster. That path only authenticates. Workloads need a
registry for main and context
images, and AWS bindings aren't wired to Pod Identity. Pass the
AWS.EKS.Cluster resource, or an aws-eks connection, to get the
full integration.