规划主机：
IP 节点类型
192.168.1.100 master/slave
192.168.1.101 master/slave
192.168.1.102 slave
192.168.1.103 slave
操作系统均为centos6.5
原理图：
1. 分别安装mysql （192.168.1.100~103）
1）下载repo，wget http://repo.mysql.com/mysql-community-release-el6-5.noarch.rpm
或从MySQL Yum仓库(http://dev.mysql.com/downloads/repo/yum/ )手动下载
2）安装repo：yum localinstall mysql-community-release-el6-5.noarch.rpm
可以通过下面的命令来确认这个仓库被成功添加：yum repolist enabled | grep community
3）yum install mysql-community-server
2. 配置文件（192.168.1.100~103）
注：这个配置是通用配置，配置好后，每个节点都可以是主节点，也可以是从节点，关键是是否启用
vim /etc/my.cnf
[client]
socket=/root/data/mysql/mysql.sock
[mysqld]
datadir=/root/data/mysql #修改默认datadir
socket=/root/data/mysql/mysql.sock
server_id=100 #每台主机不一样，根据实际情况填写唯一ID
log_bin=mysql_bin_log
binlog_format=row
relay_log = mysql_relay_bin_log
log_slave_updates = 1
read_only = 1
slave-skip-errors = 1062,1032,1060 #skip some errors such as keys conflict,data's not exists,table already exists
修改默认datadir：
vim /etc/init.d/mysqld
修改这一行为：get_mysql_option mysqld datadir "/root/data/mysql"
建立数据保存目录并赋予读写权限
chmod 755 /root
mkdir -p /root/data/mysql
chown -R mysql:mysql /root/data/mysql
chgrp -R mysql /root/data/mysql
chmod -R 755 /root/data/
3.初始化表并启动mysql
mysql_install_db --user=mysql
service mysqld start
4.账号配置
在master主机（192.168.1.100/101）上配置replication专用账号 repl，并授权
create user 'repl'@'10.%' identified by 'repl';
GRANT REPLICATION SLAVE ON *.* TO 'repl'@'10.%' IDENTIFIED BY 'repl';
注意：建议在slave上也相同的配置repl用户，方便主从切换
设置主节点(192.168.100/101)读写密码,
从节点不设置write密码和远程登录权限，防止使用slave查询过程中误用更新插入类写入语句导致产生不同步的数据）
create user 'write'@'%' identified by '123456';
grant all privileges on *.* to 'write'@'%' identified by '123456';
flush privileges;
create user 'read'@'%' identified by '123456';
GRANT select ON *.* TO 'read'@'%' IDENTIFIED BY '123456';
5.启动从服务器Replicate
除了192.168.1.101主机要从192.168.1.100同步外，其它全部从192.168.1.101同步（请看原理图）
stop slave;
CHANGE MASTER TO MASTER_HOST='192.168.1.101',
MASTER_USER='repl',
MASTER_PASSWORD='repl',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_USER='repl',
MASTER_PASSWORD='repl',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=0;
start slave;
查看从节点状态：show slave status g
第一次配置一般都会出现错误1：
错误1：Got fatal error 1236 from master when reading data from binary log: 'Could not find first log file name in b
在master那边，执行：flush logs; show master status;
+----------------------+------------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+----------------------+------------+--------------+------------------+-------------------+
| mysql_bin_log.000002 | 120| | | |
+----------------------+------------+--------------+------------------+-------------------+
记下File, Position。在slave端，执行：
stop slave;
CHANGE MASTER TO MASTER_LOG_FILE='mysql_bin_log.000002',MASTER_LOG_POS=120;
start slave;
查看状态：show slave status g
成功：
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.1.101
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql_bin_log.000006
Read_Master_Log_Pos: 120
Relay_Log_File: mysql_relay_bin_log.000002
Relay_Log_Pos: 287
Relay_Master_Log_File: mysql_bin_log.000006
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
.....
6.如何跳过一个错误？
错误2： Last_SQL_Errno: 1418
Last_SQL_Error: Error 'This function has none of DETERMINISTIC, NO SQL, or READS SQL DATA in its declaration and binary logging is enabled (you *might* want to use the less safe log_bin_trust_function_creators variable)' on query.
解决方法一：
处理临时错误，主动跳过一个错误的事务（这是我在新建一个函数是出现的，其它情况可以类似处理）
stop slave;
set GLOBAL SQL_SLAVE_SKIP_COUNTER=1;
start slave;
set GLOBAL SQL_SLAVE_SKIP_COUNTER=1;
start slave;
如果此类错较多，则在配置中忽略这种错误，
vim etc/my.cnf
在slave-skip-errors后面加 1418：
slave-skip-errors = .. ，1418 #跳过1418错误
重启mysql： serveice mysqld restart;
解决方法二：
show variables like 'log_bin_trust_function_creators';
发现是off状态，
配置文件中增加一行
log_bin_trust_function_creators=1
7.注意事项
建表时不要以自增字段作为主键，否则双机相互同步的时候难免会出现主键冲突，导致同步失败或数据丢失，建议使用guid（uuid）作为主键