How to reproduce ingress nginx issue 9222

1. Setup a cluser with 2 nodes. We used Ubuntu 18.04 with Cilium CNI but don't know if that is important.
2. helm upgrade --install ingress-nginx ingress-nginx --repo https://kubernetes.github.io/ingress-nginx --namespace ext-auth --create-namespace --version 4.1.4 -f ./ingress-nginx-values.yaml
3. kubectl apply -f ./ext-auth.yaml
4. Perform GET requests to a specific node (for example GET http) while deleting the CoreDNS pods (kubectl --namespace kube-system delete pods -l k8s-app=kube-dns)
5. Performing this a couple of times causes the bug to happen