k8s问题排查：kubernetes-dashboard crashloopbackoff 状态

现象

[root@k18-5 ~]# kubectl get pod -n kubernetes-dashboard
NAMESPACE              NAME                                         READY   STATUS             RESTARTS   AGE
kubernetes-dashboard   dashboard-metrics-scraper-856586f554-927js   1/1     Running            0          8m49s
kubernetes-dashboard   kubernetes-dashboard-7b9b87bb74-mjns7        0/1     CrashLoopBackOff   5          8m50s

kubernetes-dashboard启动后，通过浏览器访问不了。发现dashboard过一阵就从Running变为CrashLoopBackOff状态。

查看日志

kubectl logs -f -n kubernetes-dashboard kubernetes-dashboard-7b9b87bb74-mjns7

[root@k18-5 ~]# kubectl logs -f -n kubernetes-dashboard kubernetes-dashboard-7b9b87bb74-mjns7
2023/07/30 09:10:00 Starting overwatch
2023/07/30 09:10:00 Using namespace: kubernetes-dashboard
2023/07/30 09:10:00 Using in-cluster config to connect to apiserver
2023/07/30 09:10:00 Using secret token for csrf signing
2023/07/30 09:10:00 Initializing csrf token from kubernetes-dashboard-csrf secret
panic: Get "https://10.20.0.1:443/api/v1/namespaces/kubernetes-dashboard/secrets/kubernetes-dashboard-csrf": dial tcp 10.20.0.1:443: i/o timeout

goroutine 1 [running]:
github.com/kubernetes/dashboard/src/app/backend/client/csrf.(*csrfTokenManager).init(0xc00017b540)
	/home/runner/work/dashboard/dashboard/src/app/backend/client/csrf/manager.go:41 +0x413
github.com/kubernetes/dashboard/src/app/backend/client/csrf.NewCsrfTokenManager(...)
	/home/runner/work/dashboard/dashboard/src/app/backend/client/csrf/manager.go:66
github.com/kubernetes/dashboard/src/app/backend/client.(*clientManager).initCSRFKey(0xc000379300)
	/home/runner/work/dashboard/dashboard/src/app/backend/client/manager.go:502 +0xc6
github.com/kubernetes/dashboard/src/app/backend/client.(*clientManager).init(0xc000379300)
	/home/runner/work/dashboard/dashboard/src/app/backend/client/manager.go:470 +0x47
github.com/kubernetes/dashboard/src/app/backend/client.NewClientManager(...)
	/home/runner/work/dashboard/dashboard/src/app/backend/client/manager.go:551
main.main()
	/home/runner/work/dashboard/dashboard/src/app/backend/dashboard.go:95 +0x21c

根据日志分析，dashboard部署到了非master节点，网络原因无法正常运行。需要把kubernetes-dashboard部署到master节点，修改dashboard.yaml配置文件添加nodeName: master节点名，如下：

删除dashboard重新创建后查看结果

kubectl delete -f dashboard.yaml
kubectl apply -f dashboard.yaml

[root@k18-5 ~]# kubectl get pod -n kubernetes-dashboard
NAME                                         READY   STATUS    RESTARTS   AGE
dashboard-metrics-scraper-568998bfdd-4q4hh   1/1     Running   0          125m
kubernetes-dashboard-76bc846b7b-52v55        1/1     Running   0          125m

[root@k18-5 ~]# kubectl get svc -A
NAMESPACE              NAME                        TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)                  AGE
kubernetes-dashboard   dashboard-metrics-scraper   ClusterIP   10.20.89.131           8000/TCP                 133m
kubernetes-dashboard   kubernetes-dashboard        ClusterIP   10.20.24.98            443/TCP                  133m

访问 https://10.20.24.98/，成功

参考：K8S 安装kubernetes-dashboard 报错，web无法访问，查看日志报错