MHA realizes the main fault of MySQL master-slave switching

This article mainly simulates the implementation of master master failure by manually specifying master to from node.

#masterha_master_switch--master_state=dead--conf=/etc/masterha/app1.cnf--dead_master_host=10.1.1.231--dead_ master_ip=10.1.1.231--dead_master_port=63306--new_master_host=10.1.1.234--new_master_ip=10.1.1.234--new_master _port=63306

Only the following log is displayed, with some information in the attachment:

Use MHA to achieve MySQL master-slave switching main failure-Manually specify master to the photos from the node

-----Failover-----

App1:mysql Master failover 10.1.1.231 to 10.1.1.234 succeeded

Master 10.1.1.231 is down!

Check MHA Manager logs at compute-0-52.local for details.

Started manual (interactive) failover.

The latest slave 10.1.1.234 (10.1.1.234:63306) has all relay to logs for recovery.

Selected 10.1.1.234 as a new master.

10.1.1.234:ok:applying all logs succeeded.

Generating relay diff files from the latest slave succeeded.

10.1.1.234:resetting Slave info succeeded.

Master failover to 10.1.1.234 (10.1.1.234:63306) completed successfully.

At this point the library has become the main library, and the following information is viewed on the 10.1.1.234:

Mysql> Show master status;

+------------------------+----------+--------------+--------------------------------------+-------------------+

| File | Position | binlog_do_db | binlog_ignore_db | Executed_gtid_set |

+------------------------+----------+--------------+--------------------------------------+-------------------+

|      mysql-slave-bin.000011 | 120 | Denovo_ng |                   Mysql,denovo,test,information_schema | |

+------------------------+----------+--------------+--------------------------------------+-------------------+

1 row in Set (0.00 sec)

Mysql>

After the original repository is restored, add it as slave to the Master/slave cluster, as follows:

1, 10.1.1.234 (the new main library) to perform as follows:

can refer to:

centos5.6 under mysql5.6 Master-slave environment installation configuration http://www.diyoms.com/website/1318.html
　
Grant replication slave on *.* to ' Jpsync ' @ ' 10.1.1.231 ' identified by ' jppasswd ';
Flush privileges;
Also modify/ETC/MY.CNF to add the following:

　
Skip_slave_start
Note: Prevent restart of the database, start the slave process, resulting in inconsistent data.

and delete Master.info and Relay-log.info

2, 10.1.1.231 (new from the library) on the implementation:

　
Change Master to master_host= ' 10.1.1.234 ', master_port=63306,master_user= ' Jpsync ',
Master_password= ' jppasswd ', master_log_file= ' mysql-master-bin.000011 ', master_log_pos=120;
This time again to perform show slave STATUSG, the following results showed that master-slave replication has been restored.

Slave_io_running:yes

Slave_sql_running:yes