启动mysql集群报错问题

问题：

在node1、node2、node3这3个节点上部署了一个mysql集群，因为突发事件3台服务器全挂。后来服务器重启后，启动mysql服务。

操作：

我对mysql集群基本上是个小白。所以上来就到3个node上分别执行systemctl restart mysqld命令，

但是报如下错误：

Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [Note] WSREP: view((empty))
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] WSREP: failed to open gcomm backend connection: 110: failed to reach primary view: 110 (Connection timed out)
Apr 21 14:52:53 controller1 mysqld[37593]: at gcomm/src/pc.cpp:connect():158
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] WSREP: gcs/src/gcs_core.cpp:gcs_core_open():209: Failed to open backend connection: -110 (Connection timed out)
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] WSREP: gcs/src/gcs.cpp:gcs_open():1458: Failed to open channel 'my_wsrep_cluster' at 'gcomm://XXX.XXX.XXX.XXX:4567,XXX.XXX.XXX.XXX:4567,XXX.XXX.XXX.XXX:4567': -110 (Con
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] WSREP: gcs connect failed: Connection timed out
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] WSREP: wsrep::connect(gcomm://XXX.XXX.XXX.XXX:4567,XXX.XXX.XXX.XXX:4567,XXX.XXX.XXX.XXX:4567) failed: 7
Apr 21 14:52:53 controller1 mysqld[37593]: 2020-04-21 14:52:53 0 [ERROR] Aborting

 

解决办法：

在Stack Overflow上找到类似问题，按照操作，解决！

https://stackoverflow.com/questions/51414401/galera-cluster-for-multi-master-second-node-failed-to-start-in-ubuntu-16-04-4-x?r=SearchResults

简单来说就是：在node1上执行galera_new_cluster命令，启动一个新集群，然后到另外2个node上再分别执行systemctl restart mysqld命令就OK了。

上边之所以报错，是因为我启动命令用错了。