[BUG] Yurthub keeps crashing: "lb failed to cache req yurthub watch services"

[Gillefranc]

What happened:
After adding one node using the yurtadm join command, yurthub keeps crashing and restarting every few minutes, causing kube-flannel, kube-proxy and raven-agent to restart aswell

What you expected to happen:
For yurthub to keep running normally.

How to reproduce it (as minimally and precisely as possible):

• Installed kubernetes master node on a server with kubeadm (version 1.28.5).
• Installed flannel, on CIDR 10.245.0.0/16
• Installed OpneYURT components using helm, following this guide (https://openyurt.io/docs/installation/manually-setup)
• Installed yurtadm on an edge node (raspberry pi running ubuntu), and execured the following command:
  yurtadm join 10.0.20.230:6443 --token=xxx --node-type=edge --discovery-token-unsafe-skip-ca-verification --cri-socket=unix:///run/containerd/containerd.sock --v=5
• All components are installed, but the edge components crash after a few minutes, then restarting again, in a loop.

Anything else we need to know?:

Environment:

• OpenYurt version: v1.5.0
• Kubernetes version (use kubectl version):
  Client Version: v1.31.3
  Kustomize Version: v5.4.2
  Server Version: v1.28.15
• OS (e.g: cat /etc/os-release):
  NAME="Ubuntu"
  VERSION_ID="24.04"
  VERSION="24.04.1 LTS (Noble Numbat)"
• Kernel (e.g. uname -a):
  Linux thesis-gilleslefranc-2 6.8.0-49-generic set the server-addr before deploying the yurthub #49-Ubuntu SMP PREEMPT_DYNAMIC Mon Nov 4 02:06:24 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
• Install tools:
• Others:

YurtHub logs (last 10 lines):

2024-12-15T11:48:15.291964103Z stderr F I1215 11:48:15.291262       1 cache_manager.go:386] yurthub watch nodepools: /apis/apps.openyurt.io/v1beta1/nodepools get 0 objects(add:0/update:0/del:0)
2024-12-15T11:48:15.292577203Z stderr F E1215 11:48:15.291396       1 loadbalancer.go:391] lb failed to cache req yurthub watch nodepools: https://10.0.20.230:6443/apis/apps.openyurt.io/v1beta1/nodepools?allowWatchBookmarks=true&resourceVersion=5434&timeoutSeconds=309&watch=true in local cache, EOF
2024-12-15T11:48:15.292674924Z stderr F I1215 11:48:15.291069       1 util.go:248] yurthub watch nodepools: /apis/apps.openyurt.io/v1beta1/nodepools?allowWatchBookmarks=true&resourceVersion=5434&timeoutSeconds=309&watch=true with status code 200, spent 1m15.187555851s
2024-12-15T11:48:15.292730497Z stderr F I1215 11:48:15.291838       1 util.go:248] yurthub watch services: /api/v1/services?allowWatchBookmarks=true&resourceVersion=5432&timeout=5m7s&timeoutSeconds=307&watch=true with status code 200, spent 1m15.083990561s
2024-12-15T11:48:15.292761737Z stderr F I1215 11:48:15.292107       1 secure_serving.go:311] Stopped listening on 169.254.2.1:10261
2024-12-15T11:48:15.292789329Z stderr F I1215 11:48:15.287226       1 dynamic_cafile_content.go:170] "Shutting down controller" name="client-ca-bundle::/var/lib/yurthub/pki/ca.crt"
2024-12-15T11:48:15.292816385Z stderr F E1215 11:48:15.292258       1 cache_manager.go:392] yurthub /api/v1/namespaces/kube-system/configmaps watch decode ended with: EOF
2024-12-15T11:48:15.292843088Z stderr F I1215 11:48:15.287318       1 tlsconfig.go:255] "Shutting down DynamicServingCertificateController"
2024-12-15T11:48:15.292869106Z stderr F I1215 11:48:15.292326       1 cache_manager.go:386] yurthub watch configmaps: /api/v1/namespaces/kube-system/configmaps get 0 objects(add:0/update:0/del:0)
2024-12-15T11:48:15.293191137Z stderr F I1215 11:48:15.287407       1 start.go:207] hub agent exited

/kind bug

[rambohe-ch]

@Gillefranc It seems that Yurthub container has received SIGTERM signal from the os, this means that Yurthub process has been killed by os.

[rambohe-ch]

@Gillefranc you can use the following command to verify which signal cause yurthub exit.

strace -p {PID} -e trace=signal

please use the pid of yurthub process to replace the {PID}.