We upgraded our k8s worker nodes to EKS 1.19 and it broke the networking. Even though pods were running, Ingress-nginx controller stopped responding to any request. We're using Cilium 1.9.4 + Aws VPC CNI 1.7.8 combination. Looks like both Cilium & Aws VPC CNI supports 1.19.
We reverted back to 1.18 worker nodes for now, until we investigate further.
We upgraded our k8s worker nodes to EKS 1.19 and it broke the networking. Even though pods were running, Ingress-nginx controller stopped responding to any request. We're using Cilium 1.9.4 + Aws VPC CNI 1.7.8 combination. Looks like both Cilium & Aws VPC CNI supports 1.19. We reverted back to 1.18 worker nodes for now, until we investigate further.