EKS Service Controller

[jicowan]

New ACK Service Controller

Support for Amazon EKS

List of API resources

List the API resources in order of importance to you:

• Cluster
• Nodegroup
• FargateProfile

[mhausenblas]

I'm glad to see https://github.com/jicowan/fargate-operator happening and would be interested in feedback, as you roll it out and deploy it @jicowan.

[jlpettersson]

How is using this different from using AWS provider for Kubernetes Cluster API?

[rothgar]

@jlpettersson Cluster API does a lot of really cool things and is designed to be a generic way to create Kubernetes clusters. It makes some different design decisions with that goal in mind. Some include

• Cluster API is treated as your source of truth for all infrastructure. This means things like the cluster autoscaler need to be configured to use cluster api instead of AWS ASGs.
• Generic Kubernetes clusters rely on running more services in the cluster and not relying on integrations from the cloud provider. Things like metrics and logging will probably need to run inside Kubernetes instead of using services like CloudWatch
• IAM permission for controllers like CAPA will need to be more broad than EKS controller because CAPA is responsible for setting up everything needed for the cluster (VPC, gateway, etc)

With an EKS ACK controller you will get all of the configuration flexibility of the EKS api (including things like managed node groups and fargate) with limited IAM scope. You also don't need to run all of the other controllers if all you want is a way to provision an EKS cluster.

I hope that clears up some of the confusion on how the two are similar and different.

[mhausenblas]

Nice explanation Justin, thanks. Care to turn it into an FAQ item?

[agill17]

I would love to have fargate controller natively available in ACK, but in the mean time to manage dynamic stateless workload use cases I have, I ended up creating this: https://github.com/agill17/eks-fargate-controller

[ack-bot]

Issues go stale after 90d of inactivity. Mark the issue as fresh with /remove-lifecycle stale. Stale issues rot after an additional 30d of inactivity and eventually close. If this issue is safe to close now please do so with /close. Provide feedback via https://github.com/aws-controllers-k8s/community. /lifecycle stale

[RedbackThomson]

/lifecycle frozen

[elgalu]

/remove-lifecycle stale

[mikestef9]

Closing as this service controller has graduated to GA. Separate issues can be opened to discuss specific follow on topics on the controller.