54 CHEN

多台slave的mysql Replication的配置

 5.1.24版本的配置请看:http://www.masalife.com/archives/173

一、先修改服务器的配置文件

       1、Master服务器配置简单,修改my.cnf为:

       server-id       = 1

log-bin

set-variable=binlog-ignore-db=mysql

2、slave1的配置加入

server-id       = 2

master-host = 172.16.20.135

master-user = rep

master-password = cnrep

master-port = 3306

log-bin

set-variable=replicate-ignore-db=mysql

set-variable=replicate-do-db=AliSMS

set-variable=replicate-do-db=lcd

set-variable=replicate-do-db=loginmanager

set-variable=replicate-do-db=samis

set-variable=replicate-do-db=sareport

set-variable=replicate-do-db=syslog

set-variable=replicate-do-db=web_speed

log-slave-updates

3、slave2服务器的配置

server-id       = 3

master-host = 172.16.20.3

master-user = rep1

master-password = cnrep

master-port = 3306

set-variable=replicate-ignore-db=mysql

set-variable=replicate-do-db=AliSMS

set-variable=replicate-do-db=lcd

set-variable=replicate-do-db=loginmanager

set-variable=replicate-do-db=samis

set-variable=replicate-do-db=sareport

set-variable=replicate-do-db=syslog

set-variable=replicate-do-db=web_speed

二、重启master数据库

三、然后锁定master数据库的表:

mysql>FLUSH TABLES WITH READ LOCK;

四、在master数据库中添加用于slave1同步的用户,并赋予相关权限:

mysql>GRANT REPLICATION SLAVE ON *.* TO rep@sa_cfengine1 IDENTIFIED BY ‘cnrep’;

mysql>GRANT FILE,SELECT,REPLICATION SLAVE ON *.* TO rep@sa_cfengine1 IDENTIFIED BY ‘cnrep’;

五、在slave1数据库中添加用于slave2同步的用户,并赋予相关权限:

mysql>GRANT REPLICATION SLAVE ON *.* TO rep1@sa_cfengine2 IDENTIFIED BY ‘cnrep’;

mysql>GRANT FILE,SELECT,REPLICATION SLAVE ON *.* TO rep1@sa_cfengine2 IDENTIFIED BY ‘cnrep’;

六、同步数据库:

方法很多,可以打包之后scp,再解压,由于sa_cfengine1到mysql master服务器通道打通了,切sa_cfengine2到sa_cfengine1通道也打了,故直接scp整个数据库目录即可。

注意:此时要注意删除同步过来的日志文件,最好把与数据库无关的文件全删除(可以将非目录的文件全删了)。

七、重启salve1的mysql,起来之后锁定表

八、重启slave2的mysql,然后先后给slave1和master服务器的mysql表解锁

mysql> UNLOCK TABLES;

九、分别登录slave1和slave2的mysql,查看同步状态:

        mysql>SHOW SLAVE STATUS\G

*************************** 1. row ***************************

             Slave_IO_State: Waiting for master to send event

                Master_Host: 172.16.20.135

                Master_User: rep

                Master_Port: 3306

              Connect_Retry: 60

            Master_Log_File: mysql-bin.000051

        Read_Master_Log_Pos: 13856842

             Relay_Log_File: sa_cfengine1-relay-bin.000013

              Relay_Log_Pos: 624419

      Relay_Master_Log_File: mysql-bin.000051

** Slave_IO_Running: Yes**

Slave_SQL_Running: Yes

            Replicate_Do_DB: AliSMS,lcd,loginmanager,samis,sareport,syslog,web_speed

        Replicate_Ignore_DB: mysql,mysql

         Replicate_Do_Table:

     Replicate_Ignore_Table:

    Replicate_Wild_Do_Table:

Replicate_Wild_Ignore_Table:

                 Last_Errno: 0

                 Last_Error:

               Skip_Counter: 0

        Exec_Master_Log_Pos: 13856842

            Relay_Log_Space: 624419

            Until_Condition: None

             Until_Log_File:

              Until_Log_Pos: 0

         Master_SSL_Allowed: No

         Master_SSL_CA_File:

         Master_SSL_CA_Path:

            Master_SSL_Cert:

          Master_SSL_Cipher:

             Master_SSL_Key:

      Seconds_Behind_Master: 0

1 row in set (0.01 sec)

注意标注为红色的地方,两个都是yes说明一切正常,否则要检查原因,可以看error log查找原因后做相应的处理。

十、测试:

       在master数据库中update在同步列表中的一个表的一个字段,如果slave服务器的做相应改变,则测试用过。

原创文章如转载,请注明:转载自五四陈科学院[http://www.54chen.com]

Posted by 54chen linux

« yahoo、taobao云计算利器之“云”端的小飞象—Hadoop 如何解决时间同步引起oracle 不稳定的问题 »