Failed to receive reply from UDP server xxx:timeout.

[DunZane]

What happened?

i try to **https::8443** but got error like this

What did you expect to happen?

Access to dashboard WebGUI and login success.

How can we reproduce it (as minimally and precisely as possible)?

I try the step provided in github like:

1. install

helm upgrade --install kubernetes-dashboard kubernetes-dashboard/kubernetes-dashboard --create-namespace --namespace kubernetes-dashboard

2. check install

kubectl get svc -A
root@cluster-master-gz:~# kubectl get svc -A
NAMESPACE              NAME                                   TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                         AGE
default                kubernetes                             ClusterIP   10.96.0.1        <none>        443/TCP                         42h
kube-system            kube-dns                               ClusterIP   10.96.0.10       <none>        53/UDP,53/TCP,9153/TCP          42h
kubernetes-dashboard   kubernetes-dashboard-api               ClusterIP   10.104.105.185   <none>        8000/TCP                        57s
kubernetes-dashboard   kubernetes-dashboard-auth              ClusterIP   10.108.210.59    <none>        8000/TCP                        57s
kubernetes-dashboard   kubernetes-dashboard-kong-manager      NodePort    10.99.244.5      <none>        8002:31623/TCP,8445:32653/TCP   57s
kubernetes-dashboard   kubernetes-dashboard-kong-proxy        ClusterIP   10.100.99.125    <none>        443/TCP                         57s
kubernetes-dashboard   kubernetes-dashboard-metrics-scraper   ClusterIP   10.104.163.73    <none>        8000/TCP                        57s
kubernetes-dashboard   kubernetes-dashboard-web               ClusterIP   10.99.68.227     <none>        8000/TCP                        57s

3. Change kong-proxy's CLUSTER-IP from ClusterIP to NodePort

root@cluster-master-gz:~# kubectl patch service kubernetes-dashboard-kong-proxy -n kubernetes-dashboard -p '{"spec":{"type":"NodePort"}}'
service/kubernetes-dashboard-kong-proxy patched
root@cluster-master-gz:~# kubectl get svc -A
NAMESPACE              NAME                                   TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                         AGE
default                kubernetes                             ClusterIP   10.96.0.1        <none>        443/TCP                         42h
kube-system            kube-dns                               ClusterIP   10.96.0.10       <none>        53/UDP,53/TCP,9153/TCP          42h
kubernetes-dashboard   kubernetes-dashboard-api               ClusterIP   10.104.105.185   <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-auth              ClusterIP   10.108.210.59    <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-kong-manager      NodePort    10.99.244.5      <none>        8002:31623/TCP,8445:32653/TCP   2m59s
kubernetes-dashboard   kubernetes-dashboard-kong-proxy        NodePort    10.100.99.125    <none>        443:31480/TCP                   2m59s
kubernetes-dashboard   kubernetes-dashboard-metrics-scraper   ClusterIP   10.104.163.73    <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-web               ClusterIP   10.99.68.227     <none>        8000/TCP                        2m59s

4.Accssing to Dashboard(The required ports are already open on the firewall of the cloud server)

 kubectl -n kubernetes-dashboard port-forward svc/kubernetes-dashboard-kong-proxy 8443:443  --address='<sever_intranet_ip>'

5. Visit the web

https://<server_public_ip>:8443

6. Error Ocurred.
   Error
   failed the initial dns/balancer resolve for 'kubernetes-dashboard-web' with: failed to receive reply from UDP server 10.96.0.10:53: timeout.

request_id: 2172230fe938b220d29cec488bdd4a3f

Anything else we need to know?

I try the:

 kubectl -n kubernetes-dashboard port-forward svc/kubernetes-dashboard-web 8080:8000 --address='10.1.20.17'

try http://<server_public_ip>:8080/ in chrome and the login web page can be success visited but got error like:

And same issue in #9066 his problem is not set the kong-proxy from Cluster-IP to NodeProxy. But from my kubectl get svc -A output:

root@cluster-master-gz:~# kubectl get svc -A
NAMESPACE              NAME                                   TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                         AGE
default                kubernetes                             ClusterIP   10.96.0.1        <none>        443/TCP                         42h
kube-system            kube-dns                               ClusterIP   10.96.0.10       <none>        53/UDP,53/TCP,9153/TCP          42h
kubernetes-dashboard   kubernetes-dashboard-api               ClusterIP   10.104.105.185   <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-auth              ClusterIP   10.108.210.59    <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-kong-manager      NodePort    10.99.244.5      <none>        8002:31623/TCP,8445:32653/TCP   2m59s
kubernetes-dashboard   kubernetes-dashboard-kong-proxy        NodePort    10.100.99.125    <none>        443:31480/TCP                   2m59s
kubernetes-dashboard   kubernetes-dashboard-metrics-scraper   ClusterIP   10.104.163.73    <none>        8000/TCP                        2m59s
kubernetes-dashboard   kubernetes-dashboard-web               ClusterIP   10.99.68.227     <none>        8000/TCP                        2m59s

kong-proxy'type has been NodePort.

What browsers are you seeing the problem on?

No response

Kubernetes Dashboard version

NAME: kubernetes-dashboard LAST DEPLOYED: Mon May 27 08:58:08 2024 NAMESPACE: kubernetes-dashboard STATUS: deployed REVISION: 1 TEST SUITE: None

Kubernetes version

Client Version: v1.28.10 Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3 Server Version: v1.28.10

Dev environment

go version go1.22.3 linux/amd64

[floreks]

This is not how you should expose applications in Kubernetes publicly. This is a configuration issue. This repo issues are only for the actual bugs, not support.

/close

[k8s-ci-robot]

@floreks: Closing this issue.

In response to this:

This is not how you should expose applications in Kubernetes publicly. This is a configuration issue. GitHub is only for the actual bugs, not support.

/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-sigs/prow repository.

[lgmg]

Sorry, I know that this is not the place for this kind of discussion, but it may help. I had the same issue, my kubernetes cni (flannel) range pod cidr was not consistent with the one at the init.

• at kubeadm init 10.85.0.0/16
• flannel : 10.244.0.0/16
  the core dns pods was created before flannel deployment which has a different subnet range than the one where the core dns was created. I deleted the all the pods replica (kubernetes will create new ones with the correct Cluster IP subnet) and dns is working fine after. But it should be cleaner to reinstall.

[LyunSs]

I got the same issue, did you solve it? @DunZane