port 443 endpoint intermittent timeout about ingress-nginx

[k8s-ci-robot]

This issue is currently awaiting triage.

If Ingress contributors determines this is a relevant issue, they will accept it by applying the triage/accepted label and provide further guidance.

The triage/accepted label can be added by org members by writing /triage accepted in a comment.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[longwuyuan]

/remove-kind bug

• Look at the template of a new bug report in detail
• Edit your issue description and answer the questions that are asked in markdown format
• Ensure to provide all the information asked as that is how a reader takes away any potential action item

RKE installs are not a release from this project so its not likely that your use case and config is tested in the CI. So if possible install as per docs of this project and try to reproduce the problem

Its even better to ask in the RKE/Rancher forums as nobody here will have tesed/used RKE install

/kind support
/triage needs-information

[longwuyuan]

/retitle port 443 endpoint intermittent timeout

[github-actions]

This is stale, but we won't close it automatically, just bare in mind the maintainers may be busy with other tasks and will reach your issue ASAP. If you have any question or request to prioritize this, please reach #ingress-nginx-dev on Kubernetes Slack.

[aslafy-z]

Hello @longwuyuan
I tried to reproduce the problem in a external eco system but was not successful. I updated the OP, it was installed with Helm using ArgoCD, RKE is the kubernetes distribution it is installed on. I've seen a similar yet not that recurrent issue on my similar GKE install.
I'm looking for tips from people using ingress-nginx in production, it becomes a pain for me at this point.

[longwuyuan]

Your best bets are ;

• kubectl get events -A -o wide
• kubectl -n ingress-nginx logs $ingress-controller-pod-name
• kubectl get ep -A

And you can tail these things with a -w or -f for logs

If you can provide a reproduce procedure, then the problem is most likely circumstantial and so unique to your use case. And without the kubectl get events -A -o wide and logs of multiple pods, there is nothing that others can analyze and base any comments on

[longwuyuan]

Since there is no traction on this issue for so long and there is not enough data pointing to a problem in the ingress-nginx controller, I will close the issue for now as there are too many lingering inactive issues in open state. Please re-open when there is data posted here that points at or shows a problem in the controller.

/close

[k8s-ci-robot]

@longwuyuan: Closing this issue.

In response to this:

Since there is no traction on this issue for so long and there is not enough data pointing to a problem in the ingress-nginx controller, I will close the issue for now as there are too many lingering inactive issues in open state. Please re-open when there is data posted here that points at or shows a problem in the controller.

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.