I am using EKS 1.21 with 3 nodes in private subnets. There are also secondary CIDR's using another private 3 subnets for IP allocation. This is not adjustable. I want to deploy atlassian crowd, but I am not able to get the LB to create. I followed steps according https://docs.aws.amazon.com/eks/latest/userguide/aws-load-balancer-controller.html. Also checked the subnets tags and other prerequisites according https://docs.aws.amazon.com/eks/latest/userguide/alb-ingress.html. I deployed AWS ALB controller with helm and 2 pods are in running state. When I deploy Ingress resource with required annotation kubernetes.io/ingress.class: alb, LB should be created automatically, but it's not. I got this error

Warning  FailedBuildModel  4m30s (x2 over 29m)  ingress  Failed build model due to RequestError: send request failed
caused by: Post "https://elasticloadbalancing.eu-central-1.amazonaws.com/": dial tcp 10.3.22.27:443: i/o timeout

There are also 3 EniConfig resources deployed because of secondary CIDR. I am not sure if this has to do something with LB not being created and some other approach should be used. Also tried setting AWS_VPC_K8S_CNI_EXTERNALSNAT to true and created inbound rule for UDP because of node to node communication bug I read about but that didn't help. Thank you upfront for the help.