Add support for capacity unit reservation for load balancers

[shraddhabang]

Description

Application Load Balancer (ALB) and Network Load Balancer (NLB) now support Load Balancer Capacity Unit (LCU) Reservation that allows you to proactively set a minimum capacity for your load balancer, complementing its existing ability to auto-scale based on your traffic pattern. Ref : https://aws.amazon.com/about-aws/whats-new/2024/11/load-balancer-capacity-unit-reservation-application-balancers/

Adding support for this feature on controller for Ingress and Services.

Customer Experience

End user will need to update the controller to the latest version containing this feature. They will also need to update the IAM policies and IngressClassParam CRDs after updating the controller version to use this feature.

Default mode

If the alb.ingress.kubernetes.io/minimum-load-balancer-capacity or service.beta.kubernetes.io/aws-load-balancer-minimum-load-balancer-capacity annotations are not specified for the ingress or service in the manifest, the capacity will be set to zero on the new load balancers by default.

Update capacity :

If the end user wants to update the capacity for ALB/NLB to 1200 capacity units, they can update it by setting the new annotation as follows

Sets the minimum capacity for the ALB

alb.ingress.kubernetes.io/minimum-load-balancer-capacity: CapacityUnits=1200

Sets the minimum capacity for the NLB

service.beta.kubernetes.io/aws-load-balancer-minimum-load-balancer-capacity: CapacityUnits=1200

Reset capacity:

If the end user wants to reset the capacity back to 0, they can either remove the new annotations or they can set the capacity on the annotations to be zero as follows

Sets the minimum capacity for the ALB

alb.ingress.kubernetes.io/minimum-load-balancer-capacity: CapacityUnits=0

Sets the minimum capacity for the NLB

service.beta.kubernetes.io/aws-load-balancer-minimum-load-balancer-capacity: CapacityUnits=0

Turn off feature flag for capacity reservation (New)

If the customer wants to disable this feature to avoid accidental capacity reservation, then they can turn this feature off through the controller flag by updating their controller deployment as follows

      containers:
      - args:
        - --cluster-name=my-cluster
        - --ingress-class=alb
        - --log-level=debug
        - --feature-gates=LBCapacityReservation=false
        image: public.ecr.aws/eks/aws-load-balancer-controller:v2.10.0
        imagePullPolicy: IfNotPresent

Increase the stabilization requeue interval(New)

If the customer wants to increase the interval to monitor the LB status, then they can set through the controller flag by updating their controller deployment as follows

      containers:
      - args:
        - --cluster-name=my-cluster
        - --ingress-class=alb
        - --log-level=debug
        - --lb-stabilization-requeue-interval=1000s
        image: public.ecr.aws/eks/aws-load-balancer-controller:v2.10.0
        imagePullPolicy: IfNotPresent

Checklist

• Added tests that cover your change (if possible)
• Added/modified documentation as required (such as the README.md, or the docs directory)
• [x ] Manually tested
• [x ] Made sure the title of the PR is a good description that can go into the release notes

BONUS POINTS checklist: complete for good vibes and maybe prizes?! 🤯

• Backfilled missing tests for code in same general area 🎉
• Refactored something and made the world a better place 🌟

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: shraddhabang
Once this PR has been reviewed and has the lgtm label, please assign johngmyers for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[shraddhabang]

Since migration of GO SDK to V2, we dont need to generate the private SDK from model anymore. You can simple download the generated GoV2 SDK code and replace references directly.

[k8s-ci-robot]

PR needs rebase.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[k8s-ci-robot]

@shraddhabang: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
pull-aws-load-balancer-controller-unit-test	5fec250	link	true	/test pull-aws-load-balancer-controller-unit-test
pull-aws-load-balancer-controller-lint	5fec250	link	true	/test pull-aws-load-balancer-controller-lint
pull-aws-load-balancer-controller-e2e-test	5fec250	link	true	/test pull-aws-load-balancer-controller-e2e-test

Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.