502 bad gateway for pod to pod communication and unknown host exception

**Describe the bug**
We are getting 502 bad gateway for pod to pod communication for ( svc.cluster.local) traafic and also unknownhost exception for egress traffic

**Steps to reproduce**
resource "helm_release" "aws_vpc_cni" {
name = "aws-vpc-cni"
repository = "https://aws.github.io/eks-charts"
chart = "aws-vpc-cni"
namespace = "kube-system"
version ="1.18.3"

}

**Expected outcome**
no 502 bad gateway for pod to pod communication and no unknownhost exception for egress traffic

**Environment**

* Chart name:"aws-vpc-cni"
* Chart version:"1.18.3"
* Kubernetes version:1.30
* Using EKS (yes/no), if so version? yes

**Additional Context**:
This issue happening interment

Start by reproducing the issue with the provided helm_release configuration for aws-vpc-cni version 1.18.3 on EKS Kubernetes 1.30. Investigate the intermittent pod-to-pod 502 responses and egress unknown-host exceptions; done means both service-cluster.local traffic and egress traffic resolve and connect reliably.

Written by the indexing model from the issue text.