kubelet不断重启：mountpoint for cpu not found

[xpendous]

文档版本
k8s v1.6.6
containerd 1.6.1
centos stream 9 ARM版本

现象描述
kubelet不断重启
journalctl -u kubelet有如下报错：
server.go:605] failed to get the kubelet's cgroup: mountpoint for cpu not found.
server.go:271] failed to run Kubelet: mountpoint for cpu not found
kubelet.service: Main process exited, code=exited, status=255/EXCEPTION
kubelet.service: Failed with result 'exit-code'.
Stopped Kubernetes Kubelet.

或者命令行直接输入kubelet有如下报错：
I0305 15:57:27.329438 16240 server.go:410] Version: v1.16.6
I0305 15:57:27.329642 16240 plugins.go:100] No cloud provider specified.
W0305 15:57:27.329660 16240 server.go:549] standalone mode, no API client
W0305 15:57:27.329702 16240 server.go:605] failed to get the kubelet's cgroup: mountpoint for cpu not found. Kubelet system container metrics may be missing.
F0305 15:57:27.331958 16240 server.go:271] failed to run Kubelet: mountpoint for cpu not found

[xpendous]

#centos stream 9 默认使用cgroup2#
[root@cs9-01 ~]# grep cgroup /etc/mtab
cgroup2 /sys/fs/cgroup cgroup2 rw,nosuid,nodev,noexec,relatime,nsdelegate,memory_recursiveprot 0 0
[root@cs9-01 ~]#
[root@cs9-01 ~]# ll /sys/fs/cgroup
total 0
-r--r--r-- 1 root root 0 Mar 5 18:14 cgroup.controllers
-rw-r--r-- 1 root root 0 Mar 5 18:22 cgroup.max.depth
-rw-r--r-- 1 root root 0 Mar 5 18:22 cgroup.max.descendants
-rw-r--r-- 1 root root 0 Mar 5 18:14 cgroup.procs
-r--r--r-- 1 root root 0 Mar 5 18:22 cgroup.stat
-rw-r--r-- 1 root root 0 Mar 5 18:14 cgroup.subtree_control
-rw-r--r-- 1 root root 0 Mar 5 18:22 cgroup.threads
-rw-r--r-- 1 root root 0 Mar 5 18:22 cpu.pressure
-r--r--r-- 1 root root 0 Mar 5 18:22 cpuset.cpus.effective
-r--r--r-- 1 root root 0 Mar 5 18:22 cpuset.mems.effective
-r--r--r-- 1 root root 0 Mar 5 18:22 cpu.stat
drwxr-xr-x 2 root root 0 Mar 5 18:14 dev-hugepages.mount
drwxr-xr-x 2 root root 0 Mar 5 18:14 dev-mqueue.mount
drwxr-xr-x 2 root root 0 Mar 5 18:14 init.scope
-rw-r--r-- 1 root root 0 Mar 5 18:22 io.pressure
-r--r--r-- 1 root root 0 Mar 5 18:22 io.stat
-r--r--r-- 1 root root 0 Mar 5 18:22 memory.numa_stat
-rw-r--r-- 1 root root 0 Mar 5 18:22 memory.pressure
-r--r--r-- 1 root root 0 Mar 5 18:22 memory.stat
-r--r--r-- 1 root root 0 Mar 5 18:22 misc.capacity
drwxr-xr-x 2 root root 0 Mar 5 18:14 sys-fs-fuse-connections.mount
drwxr-xr-x 2 root root 0 Mar 5 18:14 sys-kernel-config.mount
drwxr-xr-x 2 root root 0 Mar 5 18:14 sys-kernel-debug.mount
drwxr-xr-x 2 root root 0 Mar 5 18:14 sys-kernel-tracing.mount
drwxr-xr-x 47 root root 0 Mar 5 18:30 system.slice
drwxr-xr-x 3 root root 0 Mar 5 18:14 user.slice

[MiV1N]

这个回答可能对题主没有帮助了，我在网上直接搜索，这个页面提的问题在搜索上面排名很高，但是没有人回答，给一个我这边遇到的问题原因，希望后面有相同问题的人可以少花点时间。

原因：

Kubernetes 从 1.19 起支持 v2 版本，如需运行 1.19 之前的老版本，需要宿主机 Docker 使用 cgroup v1
https://kind.sigs.k8s.io/docs/user/known-issues/#failure-to-create-cluster-with-cgroups-v2

我宿主机是ubuntu22.04，而ubuntu自21.04版本后的版本（不包含21.04）linux内核改用了v2版本，拉起的k8s是1.17.17的版本，还是使用的cgroup v1

解决方案：

1. 切换使用cgroup v1的系统
2. 把v2的系统切换为使用cgroupv1

ubuntu@ubuntu:~$ vim /etc/default/grub
GRUB_CMDLINE_LINUX="systemd.unified_cgroup_hierarchy=0"

ubuntu@ubuntu:~$ sudo update-grub
ubuntu@ubuntu:~$ sudo reboot