此处是在公司线上部署mysql双主时所记录的文档,为安全,IP都改为内网IP。
版本信息:
#mysql -V
mysql Ver 14.14 Distrib 5.5.37, fordebian-linux-gnu (x86_64) using readline 6.2
修改my.cf配置文件,开启bin-log功能,配置server-id。
#cat /etc/mysql/my.cnf
[myqld]
server-id = 1
log_bin =/var/log/mysql/
slave-net-timeout = 60
#salve-net-timeout默认是3600秒,缩短时间是为了防止双YES的假象
#(事实上我已遇到,参考地址:http://www.cnblogs.com/billyxp/p/3470376.html)
如果要指定同步或不同步哪些库,可使用如下参数
#binlog-do-db=osyunweidb #需要同步的数据库名,如果有多个数据库,可重复此参数,每个数据库一行
#binlog-ignore-db=mysql #不同步mysql系统数据库
查看命令 show variables like 'log_bin'; show variables like 'server_id';
mysql> show variables like 'log_bin';
+---------------+-------+
|Variable_name | Value |
+---------------+-------+
|log_bin | ON |
+---------------+-------+
1 rowin set (0.00 sec)
mysql>show variables like 'server_id';
+---------------+-------+
|Variable_name | Value |
+---------------+-------+
|server_id | 1 |
+---------------+-------+
1 rowin set (0.00 sec)
mysql> grant replication slave on *.* to replication@'%'identified by '123456'; #授权该用户对所有表都能进行复制
mysql>flush privileges; #刷新权限
mysql>flush tables with read lock; #锁定所有表,此时数据库不能写入数据
QueryOK, 0 rows affected (0.05 sec)
mysql>show master status; #查看最新bin-log文件及位置
+------------------------+------------+-------------------+-------------------------+
|File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------------+------------+-------------------+-------------------------+
|mysql-bin.000001 | 26314 | | |
+------------------------+------------+-------------------+-------------------------+
1row in set (0.00 sec)
由于退出当前mysql登陆窗口,锁表功能就失效,需克隆一个会话进行全备。
#mysqldump-uroot -p -A -B >/tmp/mysql_bak_2014_10_30.sql.gz
看下备份数据大小,确认备份成功。
#ls -l /tmp/mysql_bak_2014_10_30.sql.gz
-rw-r--r--1 root root 339222 10月 3011:01 /tmp/mysql_bak_2014_10_30.sql.gz
mysql>unlock tables;
或直接quit退出即可。
从库开启bin-log功能后,待会在主上在配置同步,互为主从就完成了。
#cat /etc/mysql/my.cnf
[myqld]
server-id = 2
log_bin =/var/log/mysql/mysql-bin.log
slave-net-timeout = 60
#salve-net-timeout默认是3600秒,缩短时间是为了防止双YES的假象
#(参考地址:http://www.cnblogs.com/billyxp/p/3470376.html)
#/etc/init.d/mysqlrestart
查看命令 show variables like 'log_bin'; show variables like 'server_id';
mysql>show variables like 'log_bin';
+---------------+-------+
|Variable_name | Value |
+---------------+-------+
|log_bin | ON |
+---------------+-------+
1row in set (0.00 sec)
mysql>show variables like 'server_id';
+---------------+-------+
|Variable_name | Value |
+---------------+-------+
|server_id | 2 |
+---------------+-------+
1row in set (0.00 sec)
在从库上解压数据。
#gzip-d mysql_bak_2014_10_30.sql.gz
登陆mysql导入数据
mysql>source /root/mysql_bak_2014_10_30.sql
因为在从库导入全备数据时,此时主库与从库的内容是一致的,但是bin-log位置不一定一致。
mysql>show master status; #查看最新bin-log文件及位置
+------------------------+------------+-------------------+-------------------------+
|File |Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------------+------------+-------------------+-------------------------+
|mysql-bin.000003 | 2328055 | | |
+------------------------+------------+-------------------+-------------------------+
1row in set (0.00 sec)
此处binlog文件与位置状态,是主库在步骤4锁表时show master status查看的位置状态。
CHANGE MASTER TO
MASTER_HOST='10.0.0.2',
MASTER_PORT=8306,
MASTER_USER='replication',
MASTER_PASSWORD='123456',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=26314;
使用start slave开启同步功能,使用show slave status\G查看同步是否成功
mysql>start slave;
QueryOK, 0 rows affected (0.00 sec)
mysql>show slave status\G #\G不按表格输出
***************************1. row ***************************
Slave_IO_State: Waiting formaster to send event
Master_Host: 10.0.0.2
Master_User: replication
Master_Port: 8306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 136270
Relay_Log_File:mysqld-relay-bin.000002
Relay_Log_Pos: 72697
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running:Yes
Slave_SQL_Running:Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
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: 98758
Relay_Log_Space: 110366
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: 622 #查看主从同步延迟,延迟大则可能需要优化
Master_SSL_Verify_Server_Cert:No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 1
1row in set (0.00 sec)
#sql线程与IO线程都是YES,slave配置成功。
由于从库是全备导入,原先在主库上配置的复制帐户也同样导入,所以这里不用在从库上新授权复制用户。
从库上的binlog文件与位置状态,是从库在刚导入时show master status查看到的位置状态。
CHANGEMASTER TO
MASTER_HOST='172.16.0.2',
MASTER_PORT=3306,
MASTER_USER='replication',
MASTER_PASSWORD='123456',
MASTER_LOG_FILE='mysql-bin.000003',
MASTER_LOG_POS=2328055;
#修改相应信息,直接把这些配置在mysql中粘贴即可。
使用start slave开启同步功能,使用show slave status\G查看同步是否成功
mysql>start slave;
QueryOK, 0 rows affected (0.00 sec)
mysql>show slave status\G
***************************1. row ***************************
Slave_IO_State: Waiting formaster to send event
Master_Host: 172.16.0.2
Master_User: replication
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000007
Read_Master_Log_Pos: 107
Relay_Log_File:mysqld-relay-bin.000006
Relay_Log_Pos: 253
Relay_Master_Log_File: mysql-bin.000007
Slave_IO_Running: Yes
Slave_SQL_Running:Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
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: 107
Relay_Log_Space: 556
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
Master_SSL_Verify_Server_Cert:No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 2
1row in set (0.00 sec)
#IO线程与sql线程都是正常。
在两台mysql各创建一个库,看两边是否都能进行同步。
分别在主库上执行 create database test01;
从库上执行create database test02;
看两台数据库上执行show databases;,看是否都有test01表和test02表。
我的经过测试,双主测试成功。
原文转自:http://yangrong.blog.51cto.com/6945369/1569700