一、背景

安裝k8s集群，初始化master節點一直不成功。

二、排查

查看日志
ps -ef | grep kube-apiserver

[root@zjy01 home]# ps -ef | grep kube-apiserver
root      85663   4637  0 12:41 pts/1    00:00:00 grep --color=auto kube-apiserver
[root@zjy01 home]# systemctl status kubelet
● kubelet.service - kubelet: The Kubernetes Node AgentLoaded: loaded (/usr/lib/systemd/system/kubelet.service; enabled; vendor preset: disabled)Drop-In: /usr/lib/systemd/system/kubelet.service.d└─10-kubeadm.confActive: activating (auto-restart) (Result: exit-code) since Mon 2025-06-30 12:42:02 CST; 5s agoDocs: https://kubernetes.io/docs/Process: 85670 ExecStart=/usr/bin/kubelet $KUBELET_KUBECONFIG_ARGS $KUBELET_CONFIG_ARGS $KUBELET_KUBEADM_ARGS $KUBELET_EXTRA_ARGS (code=exited, status=1/FAILURE)Main PID: 85670 (code=exited, status=1/FAILURE)Jun 30 12:42:02 zjy01 systemd[1]: Unit kubelet.service entered failed state.
Jun 30 12:42:02 zjy01 systemd[1]: kubelet.service failed.

journalctl -u kubelet -f

Jun 30 12:43:05 zjy01 kubelet[86386]: E0630 12:43:05.825806   86386 server.go:294] "Failed to run kubelet" err="failed to run Kubelet: misconfiguration: kubelet cgroup driver: \"systemd\" is different from docker cgroup driver: \"cgroupfs\""
Jun 30 12:43:05 zjy01 systemd[1]: Unit kubelet.service entered failed state.
Jun 30 12:43:05 zjy01 systemd[1]: kubelet.service failed.

發現是說kubelet的cgroup driver是systemd和docker的cgroupfs不一致。

三、解決

3.1 修改docker的cgroup driver為systemd

編輯 /etc/docker/daemon.json文件，添加"exec-opts": ["native.cgroupdriver=systemd"],如下：

{"exec-opts": ["native.cgroupdriver=systemd"],"data-root": "/home/docker"
}

修改后，重啟docker

systemctl daemon-reload
systemctl restart docker

查看修改結果：OK

[root@wangdy home]# docker info | grep -i cgroupCgroup Driver: systemdCgroup Version: 1

查看kubelet的cgroup driver，發現兩者已經保持一致

[root@wangdy home]# cat /var/lib/kubelet/config.yaml | grep cgroupDriver
cgroupDriver: systemd

3.2 查看結果

1）kubectl已經Active: active (running)

[root@wangdy home]# systemctl status kubelet
● kubelet.service - kubelet: The Kubernetes Node AgentLoaded: loaded (/usr/lib/systemd/system/kubelet.service; enabled; vendor preset: disabled)Drop-In: /usr/lib/systemd/system/kubelet.service.d└─10-kubeadm.confActive: active (running) since Mon 2025-06-30 12:50:17 CST; 8min agoDocs: https://kubernetes.io/docs/Main PID: 93176 (kubelet)Tasks: 63Memory: 69.7MCGroup: /system.slice/kubelet.service└─93176 /usr/bin/kubelet --bootstrap-kubeconfig=/etc/kubernetes/bootstrap-kubelet.conf --kubeconfig=/etc/kubernetes/kubelet.conf --config=/var/lib/kubelet/config...

2）pods狀態已經正常

[root@zjy01 home]# kubectl get pods -n kube-system
NAME                            READY   STATUS    RESTARTS   AGE
etcd-zjy01                      1/1     Running   0          8m53s
kube-apiserver-zjy01            1/1     Running   0          8m47s
kube-controller-manager-zjy01   1/1     Running   0          8m51s
kube-scheduler-zjy01            1/1     Running   0          8m54s
[root@zjy01 home]# 

END