二进制部署K8s集群第24节平滑切换版本技巧
作者：互联网
一、K8s平滑切换版本技巧
1.1 升级方法
1 一台服务器操作完成切换，再操作第二台、第三台...
2、删除一台node节点
3、关闭升级机器负载匀衡
4、删除老版本软链接，新版本做软链接
5、拷贝老版本重启脚本证书到新软版本软件目录
6、重启kubelet、kube-scheduler、kube-apiserver、kube-proxy、kube-controller-manager
1.2 删除Node节点
hdss7-21执行
[root@hdss7-21 opt]# kubectl get nodes -o wide
NAME STATUS ROLES AGE VERSION INTERNAL-IP EXTERNAL-IP OS-IMAGE KERNEL-VERSION CONTAINER-RUNTIME
hdss7-21.host.com Ready <none> 3d7h v1.19.2 10.4.7.21 <none> CentOS Linux 7 (Core) 3.10.0-1127.el7.x86_64 docker://19.3.13
hdss7-22.host.com Ready master 3d7h v1.19.2 10.4.7.22 <none> CentOS Linux 7 (Core) 3.10.0-1127.el7.x86_64
[root@hdss7-21 opt]# kubectl delete nodes hdss7-21.host.com
node "hdss7-21.host.com" deleted
[root@hdss7-21 opt]# kubectl get nodes
NAME STATUS ROLES AGE VERSION
hdss7-22.host.com Ready master 3d7h v1.19.21.3 注释反代
hdss7-11和hdss7-12执行
[root@hdss7-11 ~]# grep -w " server 10.4.7.21:6443" /etc/nginx/nginx.conf
# server 10.4.7.21:6443 max_fails=3 fail_timeout=30s;
nginx -s reload1.4 切目录重启
hdss7-21执行
[root@hdss7-21 ~]# mkdir /opt/net_version;cd /opt/net_version;
[root@hdss7-21 new_version]# ll
总用量 434140
-rw-r--r-- 1 root root 444555407 10月 8 2020 kubernetes-server-linux-amd64.tar.gz
tar xf kubernetes-server-linux-amd64.tar.gz
[root@hdss7-21 new_version]# ll
总用量 434140
drwxr-xr-x 4 root root 79 5月 6 13:55 kubernetes
-rw-r--r-- 1 root root 444555407 10月 8 2020 kubernetes-server-linux-amd64.tar.gz
[root@hdss7-21 new_version]# mv kubernetes /opt/kubernetes-v1.15.12
[root@hdss7-21 new_version]# cd /opt/kubernetes-v1.15.12/
[root@hdss7-21 kubernetes-v1.15.12]# rm -rf kubernetes-src.tar.gz
[root@hdss7-21 kubernetes-v1.15.12]# cd server/bin
[root@hdss7-21 bin]# rm -f *.tar *_tag
[root@hdss7-21 bin]# cp -rf /opt/kubernetes-v1.19.2/server/bin/conf .
[root@hdss7-21 bin]# cp -rf /opt/kubernetes-v1.19.2/server/bin/certs .
[root@hdss7-21 bin]# cp /opt/kubernetes-v1.19.2/server/bin/kubelet.sh .
[root@hdss7-21 opt]# ln -sf /opt/kubernetes-v1.15.12 /opt/kubernetes
[root@hdss7-21 opt]# supervisorctl status kube-apiserver-7-21 kube-controller-manager-7-21 kube-proxy-7-21 kube-scheduler-7-21
kube-apiserver-7-21 STOPPED Oct 06 08:01 AM
kube-controller-manager-7-21 STOPPED Oct 06 08:02 AM
kube-proxy-7-21 STOPPED Oct 06 08:02 AM
kube-scheduler-7-21 STOPPED Oct 06 08:02 AM
[root@hdss7-21 opt]# supervisorctl start kube-apiserver-7-21 kube-controller-manager-7-21 kube-proxy-7-21 kube-scheduler-7-21 kube-kubelet-7-21
[root@hdss7-21 bin]# supervisorctl status
etcd-server-7-21 RUNNING pid 36332, uptime 2 days, 17:43:39
flanneld-7-21 RUNNING pid 130301, uptime 2 days, 12:33:31
kube-apiserver-7-21 RUNNING pid 63354, uptime 0:09:44
kube-controller-manager-7-21 RUNNING pid 63378, uptime 0:09:13
kube-kubelet-7-21 FATAL Exited too quickly (process log may have details)
kube-proxy-7-21 RUNNING pid 63398, uptime 0:08:43
kube-scheduler-7-21 RUNNING pid 63522, uptime 0:08:121.5 开启Kubelet报错
估计是关闭kubelet没关好
看启动报错日志提示删除文件再重启
Please drain this node and delete the CPU manager checkpoint file "/data/kubelet/cpu_manager_state" before restarting Kubelet.
F1006 08:07:22.483670 62266 server.go:273] failed to run Kubelet: could not initialize checkpoint manager: could not restore state from checkpoint: checkpoint is corrupted
Please drain this node and delete the CPU manager checkpoint file "/data/kubelet/cpu_manager_state" before restarting K
[root@hdss7-21 bin]# rm -rf /data/kubelet/cpu_manager_state
[root@hdss7-21 bin]# supervisorctl start kube-kubelet-7-21
[root@hdss7-21 bin]# supervisorctl status
etcd-server-7-21 RUNNING pid 36332, uptime 2 days, 17:44:39
flanneld-7-21 RUNNING pid 130301, uptime 2 days, 12:34:31
kube-apiserver-7-21 RUNNING pid 63354, uptime 0:10:44
kube-controller-manager-7-21 RUNNING pid 63378, uptime 0:10:13
kube-kubelet-7-21 RUNNING pid 65092, uptime 0:00:37
kube-proxy-7-21 RUNNING pid 63398, uptime 0:09:43
kube-scheduler-7-21 RUNNING pid 63522, uptime 0:09:12
[root@hdss7-21 bin]# kubectl get nodes
NAME STATUS ROLES AGE VERSION
hdss7-21.host.com Ready <none> <invalid> v1.15.12 # 已切换版本
hdss7-22.host.com Ready master 3d8h v1.19.21.6 开启反代
hdss7-11和hdss7-12执行
[root@hdss7-11 ~]# grep -w " server 10.4.7.21:6443" /etc/nginx/nginx.conf
server 10.4.7.21:6443 max_fails=3 fail_timeout=30s;
nginx -s reload标签：24,hdss7,kube,21,kubernetes,二进制,opt,K8s,root 来源： https://blog.51cto.com/yht1990/2540396