Mysql主主同步失败后的恢复

基础信息

主库:

• 数据库2 10.126.4.2
• 数据库3 10.126.4.3

1. 停止数据库3对外服务

防止同步过程中服务通过数据库3写入数据

$ firewall-cmd --remove-port=3306/tcp

$ firewall-cmd --add-rich-rule="rule f amily="ipv4" source address="10.126.4.2" port protocol="tcp" port="3306" accept"

$ firewall-cmd --reload

2. 备份主库

$ mysqldump -uroot -p --single-transaction --master-data=2 --no-autocommit -A >alldatas-190708.sql

记住 MASTER_LOG_FILE 和 MASTER_LOG_POS

$ head -n 30 alldatas-190708.sql

-- CHANGE MASTER TO MASTER_LOG_FILE='mysql-bin.000016', MASTER_LOG_POS=801595396;

3. 切断数据库2、3的同步

两台主机分别执行

mysql>change master to master_host=''
mysql>stop slave;
mysql>reset slave;

4. 清空数据库3

清空数据文件

$ rm /var/lib/mysql
$ mv /db/mysql /db/mysql-190708
$ mkdir /db/mysql
$ ln -s /db/mysql /var/lib/mysql
chown mysql:mysql /var/lib/mysql
chown -h mysql:mysql /var/lib/mysql

初始化

mysqld --initialize --user=mysql
grep password /var/log/mysqld.log

重置 master 和 slave

mysql> reset master;
mysql> show master status\G
mysql> stop slave;
mysql> reset slave;
mysql> show slave status\G

             File: mysql-bin.000001
         Position: 154
     Binlog_Do_DB: 
 Binlog_Ignore_DB: 
Executed_Gtid_Set: 
1 row in set (0.00 sec)

ERROR: 
No query specified

5. 导入主库信息，重新指向主库

$ mysql -uroot -p < alldatas.sql

mysql > change master to 
        -> master_host='10.126.4.2', 
        -> master_user='repl', 
        -> master_password='********', 
        -> master_log_file='binlog.???',
        -> master_log_pos=???;
mysql > start slave;
mysql > show slave status\G

6. 数据库2配置主库信息

查询数据库3的 master_log_file 和 master_log_pos

10.126.4.3

show master status;

10.126.4.2

mysql > change master to 
        -> master_host='10.126.4.3', 
        -> master_user='repl', 
        -> master_password='********', 
        -> master_log_file='binlog.???',
        -> master_log_pos=???;
mysql > show slave status\G

7. 放开数据库3对外服务

$ firewall-cmd --reload

www.tonglei.win