Issues after enabling proxy protocol with NLB and TLS termination

[f1yers]

Hi,

I'm migrating from the ingress-nginx controller and ran into an issue when enabling proxy protocol on the NLB via the annotation "service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: '*'". After doing so, all of the nodes went unhealthy. By default, the healthcheck on the NLB that points to the /healthz page. I'm not exactly sure why this endpoint isn't returning correctly with proxy protocol enabled but I found out (by comparing to my current setup) that if I switch the NLB to use TCP healthchecks against the traffic port it all works fine.

  service.beta.kubernetes.io/aws-load-balancer-healthcheck-protocol: TCP
  service.beta.kubernetes.io/aws-load-balancer-healthcheck-port: traffic-port

I have not found anything regarding similar issues so I decided to make a post here for feedback or in case anyone else stumbles into this while migrating. Maybe there is also a better way than what I've done!
Thanks!

[vepatel]

maybe https://github.com/nginx/kubernetes-ingress/blob/main/deployments/service/loadbalancer-aws-elb.yaml can be updated with this setting, although to me this looks very cloud provider spacific