为了做实验方便,我们在同一台机器上配置两个MySQL服务(开两个端口)
1、安装、配置MySQL
事先已经安装好mysql;
1
2
3
|
[root@localhost ~] # cd /usr/local/
[root@localhost local ] # cp -r mysql/ mysql_2
[root@localhost local ] # cd mysql_2/
|
初始化mysql2,如果出现两个 “OK” 并且生成/data/mysql2目录说明正确;
1
|
[root@localhost mysql_2] # ./scripts/mysql_install_db --user=mysql --datadir=/data/mysql2
|
拷贝配置文件
1
|
[root@localhost mysql_2] # cp /etc/my.cnf ./my.cnf
|
修改配置文件相关参数,更改port 以及 socket ,并增加datadir=/data/mysql2
1
2
3
4
5
|
[root@localhost mysql_2] # vi my.cnf
[mysqld] port = 3307 socket = /tmp/mysql2 .sock
datadir= /data/mysql2
|
启动:
1
|
[root@localhost mysql_2] # /usr/local/mysql_2/bin/mysqld_safe --defaults-file=/usr/local/mysql_2/my.cnf --user=mysql &
|
如果想开机启动,需加入到/etc/rc.local 里;
1
|
# echo "/usr/local/mysql_2/bin/mysqld_safe --defaults-file=/usr/local/mysql_2/my.cnf --user=mysql &" >> /etc/rc.d/rc.local |
netstat查看已经有2个mysqld服务:
1
2
3
4
5
|
[root@localhost mysql2] # netstat -nlp |grep mysqld
tcp 0 0 0.0.0.0:3306 0.0.0.0:* LISTEN 1203 /mysqld tcp 0 0 0.0.0.0:3307 0.0.0.0:* LISTEN 1744 /mysqld unix 2 [ ACC ] STREAM LISTENING 8804 1203 /mysqld /tmp/mysql .sock
unix 2 [ ACC ] STREAM LISTENING 14159 1744 /mysqld /tmp/mysql2 .sock
|
2、 配置主从准备工作
设定mysql_2 为主(master)端口3307,mysql为从(slave)端口为3306
使用sock文件登录主mysql
1
|
[root@localhost mysql2] # /usr/local/mysql_2/bin/mysql -uroot -S /tmp/mysql2.sock
|
-S 后面指定mysql的socket文件路径,这也是登陆mysql的一种方法,因为在一台服务器上跑了两个mysql端口,所以,只能用 -S 这样的方法来区分。
在主上创建测试库db1
1
2
|
mysql> create database db1; mysql> quit |
导出主的mysql库数据然后导入给db1
1
2
|
[root@localhost mysql2] # /usr/local/mysql_2/bin/mysqldump -uroot -S /tmp/mysql2.sock mysql > 123.sql
[root@localhost mysql2] # /usr/local/mysql_2/bin/mysql -uroot -S /tmp/mysql2.sock db1 < 123.sql
|
3、配置主(master)
vim /usr/local/mysql_2/my.cnf
在[mysqld]部分查看是否有以下内容,如果没有则添加:
server-id=1
log-bin=mysql-bin
两个可选参数(2选1):
binlog-do-db=db1,db2 #需要同步的库
binlog-ignore-db=db1,db2 #忽略不同步的库
binlog-do-db=需要复制的数据库名,多个数据库名,使用逗号分隔。binlog-ignore-db=不需要复制的数据库库名,多个数据库名,使用逗号分隔。
修改配置文件后,重启mysql_2
1
2
|
[root@localhost mysql2] # pid=`ps aux |grep mysql2.sock |grep -v grep |awk '{print $2}'`[root@localhost mysql2]# kill $pid
[root@localhost mysql_2] # /usr/local/mysql_2/bin/mysqld_safe --defaults-file=/usr/local/mysql_2/my.cnf --user=mysql &
|
设置root密码:
1
2
|
[root@localhost mysql2] # /usr/local/mysql_2/bin/mysqladmin -uroot -S /tmp/mysql2.sock password '123456'
[root@localhost mysql2] # /usr/local/mysql_2/bin/mysql -uroot -S /tmp/mysql2.sock -p123456
|
mysql> grant replication slave on *.* to 'repl'@'127.0.0.1' identified by '123123';
//这里的repl是为slave端设置的访问master端mysql数据的用户,密码为123123,这里的127.0.0.1为slave的ip(本次试验配置的master和slave都在本机)。
mysql> flush privileges; //刷新库,内存的数据写入磁盘;
mysql> flush tables with read lock;
//锁定数据库,此时不允许更改任何数据
1
2
3
4
5
6
|
mysql> show master status; +------------------+----------+--------------+------------------+ | File | Position | Binlog_Do_DB | Binlog_Ignore_DB | +------------------+----------+--------------+------------------+ | mysql-bin.000004 | 378 | | | +------------------+----------+--------------+------------------+ |
//查看状态,这些数据是要记录的,一会要在slave端用到;
4、 设置从(slave)
vim /etc/my.cnf #修改或增加
server-id = 2 #这个数值不能和主一样
可选参数:replicate-do-db=db1,db2
replicate-ignore-db=db1,db2 #意义同主的那两个可选参数
重启从的mysql服务:service mysqld restart
拷贝主的db1库数据到从:
先导出db1.sql文件,然后在从数据库中同样创建数据库db1,导入主的db1.sql文件到从;
1
2
3
|
[root@localhost ~] # /usr/local/mysql_2/bin/mysqldump -uroot -S /tmp/mysql2.sock -p123456 db1 > db1.sql
[root@localhost ~] # /usr/local/mysql/bin/mysql -uroot -S /tmp/mysql.sock -e "create database db1"
[root@localhost ~] # /usr/local/mysql/bin/mysql -uroot -S /tmp/mysql.sock db1 < db1.sql
|
登陆从的mysql
1
2
3
|
[root@localhost ~] # /usr/local/mysql/bin/mysql -uroot -S /tmp/mysql.sock
mysql> slave stop; mysql> change master to master_host= '127.0.0.1' ,master_port=3307,master_user= 'repl' ,master_password= '123123' ,master_log_file= 'mysql-bin.000004' ,master_log_pos=378;
|
//master_log_file='mysql-bin.000004',master_log_pos=378为主show master status显示的前2列内容;
1
|
mysql> slave start; |
主上,解锁表:
1
|
[root@localhost ~] # /usr/local/mysql_2/bin/mysql -uroot -S /tmp/mysql2.sock -p123456 -e "unlock tables"
|
登录从
1
|
[root@localhost ~] # /usr/local/mysql/bin/mysql -uroot -S /tmp/mysql.sock
|
查看从的状态 show slave status\G;
确认以下两项参数都为yes:
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
|
mysql> show slave status\G; *************************** 1. row *************************** Slave_IO_State: Waiting for master to send event
Master_Host: 127.0.0.1
Master_User: repl
Master_Port: 3307
Connect_Retry: 60
Master_Log_File: mysql-bin.000004
Read_Master_Log_Pos: 378
Relay_Log_File: localhost-relay-bin.000002
Relay_Log_Pos: 251
Relay_Master_Log_File: mysql-bin.000004
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: 378
Relay_Log_Space: 410
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:
|
5、 测试主从
主上清空db1库的db表 :
登录主
1
2
3
4
5
6
7
8
|
[root@localhost ~] # mysql -uroot -S /tmp/mysql2.sock -p123456
mysql> use db1; mysql> select count(*) from db;
+----------+ | count(*) | +----------+ | 2 | +----------+ |
清空db表;
1
2
3
4
5
6
7
|
mysql> truncate table db; mysql> select count(*) from db;
+----------+ | count(*) | +----------+ | 0 | +----------+ |
进入slave,查看db1库db表,和主的数据一样,表示同步OK;
1
2
3
4
5
6
7
8
|
[root@localhost ~] # mysql -uroot -S /tmp/mysql.sock
mysql> use db1; mysql> select count(*) from db;
+----------+ | count(*) | +----------+ | 0 | +----------+ |
登录主,删除表db;
1
2
|
[root@localhost ~] # mysql -uroot -S /tmp/mysql2.sock -p123456
mysql> drop table db; |
登录从,查看db表不存在;
1
2
3
|
[root@localhost ~] # mysql -uroot -S /tmp/mysql.sock
mysql> select * from db;
ERROR 1146 (42S02): Table 'db1.db' doesn't exist
|
建议: MySQL主从机制比较脆弱,谨慎操作。如果重启master,务必要先把slave停掉,也就是说需要在slave上去执行 slave stop 命令,然后再去重启master的mysql服务,否则很有可能就会中断了。当然重启完后,还需要把slave给开启 slave start。