Outdated documentation for Service object annotation "service.beta.kubernetes.io/aws-load-balancer-healthcheck-timeout"

[saikrishnaburugula]

Describe the bug
Service Annotations documentation states the following for annotation "service.beta.kubernetes.io/aws-load-balancer-healthcheck-timeout"

The controller currently ignores the timeout configuration due to the limitations on the AWS NLB. The default timeout for TCP is 10s and HTTP is 6s.

Reference: https://kubernetes-sigs.github.io/aws-load-balancer-controller/latest/guide/service/annotations/#healthcheck-timeout

As per AWS NLB documentation https://docs.aws.amazon.com/elasticloadbalancing/latest/network/target-group-health-checks.html#health-check-settings

HealthCheckTimeoutSeconds : The amount of time, in seconds, during which no response from a target means a failed health check. The range is 2–120 seconds. The default values are 6 seconds for HTTP and 10 seconds for TCP and HTTPS health checks.

As per the AWS load balancer controller documentation https://kubernetes-sigs.github.io/aws-load-balancer-controller/v2.8/deploy/configurations/ the feature gate NLBHealthCheckAdvancedConfiguration should be enabled or disabled
to set advanced health check configuration for NLB, for example health check timeout.

Steps to reproduce

• Using the service annotations it is possible to set the timeout value to a different value than default as specified in the NLB documentation.

Expected outcome
Update the AWS LBC documentation to reflect the latest information regarding the usage of service.beta.kubernetes.io/aws-load-balancer-healthcheck-timeout and confirm if controller currently ignores the timeout configuration due to the limitations on the AWS NLB. If yes, please update the documentation with the limitations.
Environment

• AWS Load Balancer controller version: v2.8
• Kubernetes version: 1.29
• Using EKS (yes/no), if so version? Yes. 1.29

[andreybutenko]

Thanks for the feedback! You're right, we should 1/ update the controller logic to remove this restriction on NLB health check settings, and 2/ update the controller documentation accordingly. We are open to PRs for this :)

[Shiravani24]

Would like to work on this request

[luca-rui]

Is there any soft timeline on when we can expect the custom timeouts to be supported?