树莓派4 Kubernetes集群kube-scheduler启动失败：连接拒绝问题排查

环境：基于Ubuntu 22.04.1 LTS搭建的1个控制节点+3个工作节点树莓派4 Kubernetes集群，当前优先聚焦kube-scheduler持续失败问题排查。

1. 查看集群Pod状态 #

执行命令：

peter@k8s-controller:~/.kube$ kubectl get pods --all-namespaces

返回Pod状态：

NAMESPACE      NAME                                     READY   STATUS             RESTARTS          AGE
kube-flannel   kube-flannel-ds-tsvqz                    0/1     CrashLoopBackOff   7 (62s ago)       14m
kube-system    etcd-k8s-controller                      1/1     Running            599 (11m ago)     21h
kube-system    kube-apiserver-k8s-controller            1/1     Running            593 (3m3s ago)    21h
kube-system    kube-controller-manager-k8s-controller   1/1     Running            202 (3m1s ago)    20h
kube-system    kube-scheduler-k8s-controller            0/1     CrashLoopBackOff   646 (3m23s ago)   21h

2. 分析kube-scheduler日志 #

执行命令查看日志：

peter@k8s-controller:~/.kube$ kubectl logs kube-scheduler-k8s-controller  -n kube-system

日志情况：kube-scheduler成功获取kube-system/kube-scheduler租约后，多次出现连接拒绝错误，最终因租约续约超时丢失领导者选举，导致调度队列关闭。核心错误信息：

E1110 13:25:53.900211       1 leaderelection.go:330] error retrieving resource lock kube-system/kube-scheduler: Get "https://192.168.1.135:6443/apis/coordination.k8s.io/v1/namespaces/kube-system/leases/kube-scheduler?timeout=5s": dial tcp 192.168.1.135:6443: connect: connection refused

内容的提问来源于stack exchange，提问作者peterlandis