Mysql主从配置

`sudo vi /etc/mysql/mysql.conf.d/mysqld.cnf`

# enable remote connection
bind-address = 0.0.0.0

...

# 2017-5-19 16:35
server-id = 1
log_bin = /var/log/mysql/mysql-bin.log

• 重启mysql

`sudo service mysql restart`

• 授权主从复制的用户

`grant replication slave on *.* to ‘replication‘@‘192.168.68.%‘ identified by ‘replication‘;`

• `flush tables with read lock;`

• `show master status;`

• 备份整库并传到从服务器

`mysqldump -uroot -p --all-databases > /tmp/mysql_bak.sql`

`scp /tmp/mysql_bak.sql test@192.168.68.23:/tmp/mysql_bak.sql`

• `unlock tables;`

3. 配置从服务器

• 修改mysql配置

`sudo vi /etc/mysql/mysql.conf.d/mysqld.cnf`

# enable remote connection
bind-address = 0.0.0.0

...

# 2017-5-19 16:35 该值需要与主服务器不一样
server-id = 2

• 重启mysql

`sudo service mysql restart`

• 初始化整库

`mysql -uroot -p < /tmp/mysql_bak.sql`

• 添加认从配置

mysql> CHANGE MASTER TO
-> MASTER_HOST=‘192.168.68.24‘,
-> MASTER_PORT=3306,
-> MASTER_USER=‘replication‘,
-> MASTER_PASSWORD=‘replication‘,
-> MASTER_LOG_FILE=‘mysql-bin.000001‘,
-> MASTER_LOG_POS=120;

• 启动主从

`start slave;`

• 查看主从状态

` show slave status \G`

Slave_IO_Running/Slave_SQL_Running为Yes则表示成功

4. 遇到问题

The slave I/O thread stops because master and slave have equal MySQL server UUIDs; these UUIDs must be different for replication to work?

解决方案：http://www.chriscalender.com/resolving-error-master-and-slave-have-equal-mysql-server-uuids/

a) `mv /var/lib/mysql/auto.cnf /tmp/auto.cnf`

b) `sudo service mysql restart`

Mysql主从配置

标签：scp   targe   mysq   mysql配置   pre   ice   主服务器   base   diff