使用MMM搭建Mysql同步高可用性
先介绍下MMM?
MMM是 mysql-master-master的缩写。?
MMM项目来自 Google：http://code.google.com/p/mysql-master-master?
MySQL本身没有提供replication failover的解决方案，那么如何使Replication方案具有HA的功能呢，那就是MMM。?
大家都知道mysql的同步机制在解决实际问题上具有很高的可行性，特别对读写负载比较高的web2.0来说，那么在实际应用中必须保证mysql的稳定性及性能。?
Master-Master Replication?
1、使用两个MySQL数据库db1,db2，互为Master和Slave，即：?
一边db1作为db2的master，一旦有数据写向db0时，db2定时从db1更新?
另一边db2也作为db1的master，一旦有数据写向db2时，db1也定时从db2获得更新?
2、但从AppServer的角度来说，同时只有一个结点db1扮演Master，另外一个结点db2扮演Slave，不能同时两个结点扮演Master。即AppSever总是把write操作分配某个数据库(db1)，除非db1 failed，被切换。?
3、如果扮演Slave的数据库结点db2 Failed了：?
a)此时appServer要能够把所有的read,write分配给db1，read操作不再指向db2?
b)一旦db2恢复过来后，继续充当Slave角色，并告诉AppServer可以将read分配给它了?
4、如果扮演Master的数据库结点db1 Failed了?
a)此时appServer要能够把所有的写操作从db1切换分配给db2，也就是切换Master由db2充当?
b)db1恢复过来后，充当Slave的角色，Master由db2继续扮演?
MMM利用了虚拟IP的技术：1个网卡可以同时使用多个IP。?
(所以使用MMM时，需要2*n+1个IP，n为mysql数据库结点个数，包括master,slave)?
MMM有3个重要的器件:?
1、mmmd_mon – 数据库控制节点运行。?
2、mmm_control – 数据库控制节点运行。?
3、mmmd_agent – 数据库节点（master/slave）运行。?
本文环境：?
server1 ip: 192.168.1.225 ; virtual read ip:192.168.1.229?
server2 ip: 192.168.1.226 ; virtual read ip:192.168.1.230?
server3 ip: 192.168.1.227?
virtual write ip: 192.168.1.231?
配置过程很简单，?
server1 server2 replication双向 master-master?
server1 server2 安装mmm并配置mmm_agent.conf?
server3 安装mmm并配置mmm_mon.conf?
1，首先我们要先配置好master-master双向同步，这一步就不多说了。?
2，安装部署MMM?
3台服务器都要安装MMM软件，安装过程如下：?
安装如下perl模块，为MMM安装做好准备，?
cpan Algorithm::Diff?
cpan Proc::Daemon?
cpan Time::HiRes?
cpan DBI?
cpan DBD::mysql?
wget http://mysql-master-master.googlecode.com/files/mmm-1.0-pre2.tar.bz2?
tar xzf mmm-1.0-pre2.tar.bz2?
cd mmm-1.0-pre2?
./install.pl?
3，在server1上配置MMM?
修改配置 /usr/local/mmm/etc/mmm_agent.conf，如下：?
#?
# Master-Master Manager config (agent)?
#?
# Debug mode?
debug no?
# Paths?
pid_path /usr/local/mmm/var/mmmd_agent.pid?
bin_path /usr/local/mmm/bin?
# Logging setup?
log mydebug?
file /usr/local/mmm/var/mmm-debug.log?
level debug?
log mytraps?
file /usr/local/mmm/var/mmm-traps.log?
level trap?
# MMMD command socket tcp-port and ip?
bind_port 9989?
# Cluster interface?
cluster_interface eth0?
# Define current server id?
this db1?
mode master?
# For masters?
peer db2?
# Cluster hosts addresses and access params?
host db1?
ip 192.168.1.225?
port 3306?
user rep_agent?
password RepAgent?
host db2?
ip 192.168.1.226?
port 3306?
user rep_agent?
password RepAgent?
GRANT ALL PRIVILEGES on *.* to ‘rep_agent’@’%′ identified by ‘RepAgent’;?
GRANT ALL PRIVILEGES on *.* to ‘rep_monitor’@’%’ identified by ‘RepMonitor’;
?
?
3，在server2上配置MMM?
修改配置 /usr/local/mmm/etc/mmm_agent.conf，如下：?
#?
# Master-Master Manager config (agent)?
#?
# Debug mode?
debug no?
# Paths?
pid_path /usr/local/mmm/var/mmmd_agent.pid?
bin_path /usr/local/mmm/bin?
# Logging setup?
log mydebug?
file /usr/local/mmm/var/mmm-debug.log?
level debug?
log mytraps?
file /usr/local/mmm/var/mmm-traps.log?
level trap?
# MMMD command socket tcp-port and ip?
bind_port 9989?
# Cluster interface?
cluster_interface eth0?
# Define current server id?
this db2?
mode master?
# For masters?
peer db1?
# Cluster hosts addresses and access params?
host db1?
ip 192.168.1.225?
port 3306?
user rep_agent?
password RepAgent?
host db2?
ip 192.168.1.226?
port 3306?
user rep_agent?
password RepAgent?
GRANT ALL PRIVILEGES on *.* to ‘rep_agent’@’%′ identified
