背景：k8s集群其中一个节点在人为重启之后，k8s节点的kubelet、docker 启动失败，并且服务器的文件系统变成只读权限，导致tab键不可用。直接执行重启docker命令不管用！

Jul 18 15:10:18 master dockerd[15032]: time=“2022-07-18T15:10:18.617902468+08:00” level=info msg=“Starting up”
Jul 18 15:10:18 master dockerd[15032]: time=“2022-07-18T15:10:18.620916549+08:00” level=info msg=“parsed scheme: “unix”” module=grpc
Jul 18 15:10:18 master dockerd[15032]: time=“2022-07-18T15:10:18.620940189+08:00” level=info msg=“scheme “unix” not registered, fallback to default scheme” module=grpc
Jul 18 15:10:18 master dockerd[15032]: time=“2022-07-18T15:10:18.620962547+08:00” level=info msg=“ccResolverWrapper: sending update to cc: {[{unix:///run/containerd/containerd.sock 0 }] }” module=grpc
Jul 18 15:10:18 master dockerd[15032]: time=“2022-07-18T15:10:18.620981158+08:00” level=info msg=“ClientConn switching balancer to “pick_first”” module=grpc
Jul 18 15:10:18 master dockerd[15032]: failed to start daemon: failed to dial “/run/containerd/containerd.sock”: unknown service containerd.services.namespaces.v1.Namespaces: not implemented
Jul 18 15:10:18 master systemd[1]: docker.service: Main process exited, code=exited, status=1/FAILURE

恢复文件系统，解除只读限制

mount -o remount,rw /

查看docker 状态

systemctl status docker 

启动docker

systemctl start docker 

查看日志

failed to start daemon: failed to dial “/run/containerd/containerd.sock”: unknown service containerd.services.namespaces.v1.Namespaces: not implemented

启动 containerd

systemctl start containerd

然后再重启 docker