Errors from NGINX after unrelated application deployment

[ryan-dyer-sp]

trafficstars

Describe the bug First off I've been using nginx ingress controller for years and never seen anything like this. Have been using NGF a few months now and we've started seeing a very odd behavior where one of our apps gets deployed and coincidentally the NGF starts throwing errors that never go away until the NGF controller pod is restarted. However once that happens sometimes NGF seems to want to redeploy all the gateway pods which results in all of our services having a very brief outage; even though we run multiple pods per gateway.

To Reproduce Steps to reproduce the behavior: Have not been able to come up with a reproduction scenario. We deploy applications multiple times a day without issue. But on occasion things break.

Expected behavior Application deployments dont put NGF into a bad state.

Your environment

• Version of the NGINX Gateway Fabric. - 2.1.4 - 2d999117b228356b432ccb1e07bd0ceca7b9f2ef

• Version of Kubernetes - 1.32.9-eks

• Kubernetes platform (e.g. Mini-kube or GCP) - eks

• Details on how you expose the NGINX Gateway Fabric Pod (e.g. Service of type LoadBalancer or port-forward) The gateways are exposed as AWS NLBs (LoadBalancer)

• Logs The error waiting for connection errors tend to occur every 30-40s once the configuration error one is seen. LMK if you need full logs from around the time it occurs.