版本:mysql-5.6.11
LINK:mysql环境搭建
实验:主从服务器在同一台服务器,不同服务器类似。
============================
1)修改my.cnf配置文件，开启多应用

[mysqld_multi]
mysqld     = /usr/local/mysql/bin/mysqld_safe
mysqladmin = /usr/local/mysql/bin/mysqladmin
user       = root
#password  = your_password #服务器修改密码后开启
#主服务器配置mysqld1
[mysqld1]
lower_case_table_names=1 #主从服务器采用linux-windows不同系统开启
datadir    = /usr/local/mysql/data
pid-file   = /usr/local/mysql/data/master0001.pid
port       = 3306
server_id  = 1
socket     = /tmp/mysql.sock
user       = mysql
log_slave_updates = /usr/local/mysql/log/mysql_slave_updates.log
log-error  = /usr/local/mysql/log/error.log
long_query_time = 1
slow_query_log_file = /usr/local/mysql/log/slowquery.log
log-queries-not-using-indexes = on
log-bin    = /usr/local/mysql/log/binary/mysql-bin-log
#备份的数据库
binlog-do-db = wp
binlog-do-db = phpmyadmin
#不需备份的数据库
binlog-ignore-db = mysql
binlog-ignore-db = performance_schema
#提交立刻写日志
sync_binlog=1
innodb-flush_logs_at_trx_commit=1

#从服务器mysqld2
[mysqld2]
lower_case_table_names=1
datadir    = /usr/local/mysql/data2
pid-file   = /usr/local/mysql/data2/mslave0002.pid
port       = 3307
server_id  = 2
socket     = /tmp/mysql.sock2
user       = mysql
report-user = mslave
report-host = localhost
report-port = 3307
relay-log = /usr/local/mysql/log2/relay/mysql-relay
relay-log-info-file = /usr/local/mysql/log2/relay/mysql-relay.info
log_slave_updates = /usr/local/mysql/log2/mysql_slave_updates.log
log-error  = /usr/local/mysql/log2/error.log
replicate-do-db = wp
replicate-ignore-db = phpmyadmin
replicate-ignore-db = mysql
replicate-ignore-db = performance_schema

2)创建日志文件(夹)

mkdir -p /usr/local/mysql/log/binary/touch /usr/local/mysql/log/error.logtouch /usr/local/mysql/log/mysql_slave_updates.logtouch /usr/local/mysql/log/slowquery.logmkdir -p /usr/local/mysql/log2/relay/touch /usr/local/mysql/log2/error.logtouch /usr/local/mysql/log2/mysql_slave_updates.log#从(多应用的其他服务器)服务器需要copy 主服务器data/mysql 文件至自己的目录data2/mysqlcp -r /usr/local/mysql/data/mysql /usr/local/mysql/data2/mysqlchown -R mysql:mysql /usr/local/mysql#启动命令要有--defaults-file参数，否则报错mysqld_multi --defaults-file=/etc/my.cnf start 1-2

4)启动mysql，修改访问权限

#主服务器/usr/local/mysql/bin/mysqld_multi --defaults-file=/etc/my.cnf start 1,2/usr/local/mysql/bin/mysql -u root -S /tmp/mysql.sock
 
GRANT ALL PRIVILEGES ON *.* TO "root"@"%" IDENTIFIED BY "your_password";GRANT REPLICATION SLAVE ON *.* TO "mslave"@"%" IDENTIFIED BY 'your_password2';USE mysql;UPDATE user SET Password = password('your_password') WHERE User='root';FLUSH PRIVILEGES;
 
#从服务器/usr/local/mysql/bin/mysql -u root -S /tmp/mysql.sock2
 
GRANT ALL PRIVILEGES ON *.* TO "root"@"%" IDENTIFIED BY "your_password";USE mysql;UPDATE user SET Password = password('your_password') WHERE User='root';FLUSH PRIVILEGES;

启用密码重启mysql服务(修改my.cnf)
4)配置主从同步
a.主服务器操作

FLUSH TABLES WITH READ LOCK;SHOW MASTER STATUS;+----------------------+----------+---------------+--------------------------+-------------------+| File                 | Position | Binlog_Do_DB  | Binlog_Ignore_DB         | Executed_Gtid_Set |+----------------------+----------+---------------+--------------------------+-------------------+| mysql-bin-log.000147 | 105607   | wp,phpmyadmin | mysql,performance_schema |                   |+----------------------+----------+---------------+--------------------------+-------------------+1 row in set (0.00 sec)

记录File 和 Position 值

备份&还原数据库
参考 mydumper多线程备份还原mysql

主服务器解锁

UNLOCK TABLES;

b.从服务器操作

CHANGE MASTER TO MASTER_HOST='127.0.0.1',MASTER_USER='mslave',MASTER_PASSWORD='your_password2',MASTER_LOG_FILE='mysql-bin-log.000147',MASTER_LOG_POS=105607,MASTER_PORT=3306;START SLAVE;SHOW SLAVE STATUS;/*
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
即成功，错误信息可以根据Last_IO_Errno:和Last_SQL_Error:排查。
*/

从服务器在其它实例或服务器上配置类似,仿照[mysqld2],确保server_id值不同。