http://ywliyq.blog.51cto.com/11433965/1856974
本文为南非蚂蚁的书籍《循序渐进linux-第二版》-8.4的读笔记
MMM集群套件（MYSQL主主复制管理器）
MMM套件主要的功能是通过下面三个脚本实现的
1)mmm_mond
这是一个监控进程，运行在管理节点上，主要负责都所有数据库的监控工作，同时决定和处理所有节点的角色切换
2)mmm_agentd
这是一个代理进程，运行在每个MYSQL服务器上，主要完成监控的测试工作以及执行简单的远端服务设置
3)mmm_control
简单的管理脚本，用来查看和管理集群运行状态，同事管理mmm_mond进程
MMM方案并不太适应于对数据安全性要求很高并且读、写频繁的环境中
================================================
8.4.2 MMM典型应用方案
MMM双Master节点应用架构
在双Master节点的基础上，增加多个Slave节点，即可实现双主多从节点应用架构
双主多从节点的MYSQL架构适合读查询量非常大的业务环境，通过MMM提供的读IP和写IP可以轻松实现MYSQL的读写、分离架构
================================================
8.4.3 MMM高可用MYSQL方案架构图
双主双从的MYSQL高可用集群架构
服务器配置环境如表：
MMM双主双从应用架构对应的读、写分离IP列表：
8.4.4 MMM的安装与配置
4个主从节点上使用yum安装mysql数据库并设置密码
# yum -y install mysql mysql-server
启动mysql
# /etc/init.d/mysqld start
创建mysql密码：(jzh0024)
# mysql_secure_installation
默认密码为空，一直y即可
至此，mysql数据库安装完成。
-----------------------------
1.修改mysql配置文件，所有mysql主机上设置read_only参数，
/etc/my.cnf配置,[mysqld]段添加：
server-id = 1
log-bin=mysql-bin
relay-log = mysql-relay-bin
replicate-wild-ignore-table=mysql.%
replicate-wild-ignore-table=test.%
replicate-wild-ignore-table=information_schema.%
read_only=1
其中\
Master1的server-id = 1
Master2的server-id = 2
Slave1的server-id = 3
Slave1的server-id = 4
重启mysql数据库
# /etc/init.d/mysqld restart
-----------------------------
1.MMM套件的安装
使用yum方式安装MMM套件，所有节点安装epel的yum源
# cd /server/tools/
上传epel-release-6-8.noarch.rpm源文件
# rpm -Uvh epel-release-6-8.noarch.rpm
在monitor节点执行命令：
[root@Monitor tools]# yum -y install mysql-mmm*
4个MYSQL_db节点只需要安装mysql-mmm-agent即可，执行命令
[root@Master1 tools]# yum -y install mysql-mmm-agent
安装完成后查看安装的mmm版本
[root@Monitor tools]# rpm -qa|grep mysql-mmm
mysql-mmm-2.2.1-2.el6.noarch
mysql-mmm-tools-2.2.1-2.el6.noarch
mysql-mmm-monitor-2.2.1-2.el6.noarch
mysql-mmm-agent-2.2.1-2.el6.noarch
至此，MMM集群套件安装完成。
-------------------------------------
2. MMM集群套件的配置
在进行MMM套件配置之前，需要事先配置Master1到Master2之间的主主互为同步，
同时还需要配置Master1到Slave1、Slave2之间为主从同步
一、配置Master1到Master2之间的主主互为同步
Master1先创建一个数据库及表，用于同步测试
mysql> create database master001;
mysql> use master001;
创建表
mysql> create table permaster001(member_no char(9) not null,name char(5),birthday date,exam_score tinyint,primary key(member_no));
查看表内容
mysql> desc permaster001;
Master1进行锁表并备份数据库
mysql> flush tables with read lock;
Query OK, 0 rows affected (0.00 sec)
不要退出终端,否则锁表失败;新开启一个终端对数据进行备份,或者使用mysqldump进行备份
# cd /var/lib/
# tar zcvf mysqlmaster1.tar.gz mysql
# scp -P50024 mysqlmaster1.tar.gz [email protected]:/var/lib/
[email protected]'s password:
mysqlmaster1.tar.gz 100% 214KB 213.9KB/s 00:00
注意:此处需要开启Master2授权root远程登录
# vim /etc/ssh/sshd_config
#PermitRootLogin no
并重启ssh连接
[root@Master1 lib]# /etc/init.d/sshd restart
数据传输到Master2后,依次重启Master1,Master2的数据库
[root@Master1 lib]# /etc/init.d/mysqld restart
Stopping mysqld: [ OK ]
Starting mysqld: [ OK ]
[root@Master2 tools]# /etc/init.d/mysqld restart
Stopping mysqld: [ OK ]
Starting mysqld: [ OK ]
3.创建复制用户并授权
Master1上创建复制用户,
mysql> grant replication slave on *.* to 'repl_user'@'10.24.24.21' identified by 'repl_password';
Query OK, 0 rows affected (0.00 sec)
刷新授权表
mysql> grant replication slave on *.* to 'repl_user'@'10.24.24.21' identified by 'repl_password';
Query OK, 0 rows affected (0.00 sec)
mysql> flush privileges;
Query OK, 0 rows affected (0.00 sec)
mysql> show master status;
+------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000002 | 345 | | |
+------------------+----------+--------------+------------------+
1 row in set (0.00 sec)
然后在Master2的数据库中将Master1设为自己的主服务器
# cd /var/lib/
# tar xf mysqlmaster1.tar.gz
mysql> change master to \
-> master_host='10.24.24.20',
-> master_user='repl_user',
-> master_password='repl_password',
-> master_log_file='mysql-bin.000002',
-> master_log_pos=345;
需要注意master_log_file和master_log_pos选项，这两个值是刚才在Master1上查询到的结果
Master2上启动从服务器，并查看DB2上的从服务器运行状态
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.24.24.20
Master_User: repl_user
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 345
Relay_Log_File: mysql-relay-bin.000002
Relay_Log_Pos: 251
Relay_Master_Log_File: mysql-bin.000002
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
至此，Master1到Master2的MYSQL主从复制已完成。
验证数据的完整性,自己创建库或者表来进行验证数据是否同步。
---------------------------------------------
二、配置Master2到Master1的主从复制
Master2数据库中创建复制用户
mysql> grant replication slave on *.* to 'repl_user1'@'10.24.24.20' identified by 'repl_password1';
刷新授权表
mysql> flush privileges;
Query OK, 0 rows affected (0.00 sec)
mysql> show master status;
+------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000002 | 347 | | |
+------------------+----------+--------------+------------------+
1 row in set (0.00 sec)
在Master1的数据库中将Master2设为自己的主服务器
mysql> change master to \
-> master_host='10.24.24.21',
-> master_user='repl_user1',
-> master_password='repl_password1',
-> master_log_file='mysql-bin.000002',
-> master_log_pos=347;
在Master1上启动从服务器
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
查看Master1上从服务器的运行状态
mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.24.24.21
Master_User: repl_user1
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 347
Relay_Log_File: mysql-relay-bin.000002
Relay_Log_Pos: 251
Relay_Master_Log_File: mysql-bin.000002
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: mysql.%,test.%,information_schema.%
Slave_IO_Running和Slave_SQL_Running都处于YES状态。表明DB1上复制服务运行正常，mysql双主模式主从复制配置完毕。
验证数据的完整性,自己创建库或者表来进行验证数据是否同步。
-------------------------------------------------------
三、配置Master1和Slave1、Slave2之间的主从同步
需要注意的是，在Slave1、Slave2和主Master同步时，"Master_Host"的地址要添加Master节点的物理IP地址，而不是虚拟IP地址。
Master1先创建一个数据库及表，用于同步测试
mysql> create database slave012;
mysql> use slave012;
创建表
mysql> create table persalve012(member_no char(9) not null,name char(5),birthday date,exam_score tinyint,primary key(member_no));
查看表内容
mysql> desc persalve012;
Master1进行锁表并备份数据库
mysql> flush tables with read lock;
Query OK, 0 rows affected (0.00 sec)
不要退出终端,否则锁表失败;新开启一个终端对数据进行备份,或者使用mysqldump进行备份
# cd /var/lib/
# tar zcvf mysqlslave12.tar.gz mysql
发送到Slave1:
[root@Master1 lib]# scp -P50024 mysqlslave12.tar.gz [email protected]:/var/lib/
The authenticity of host '[10.24.24.22]:50024 ([10.24.24.22]:50024)' can't be established.
RSA key fingerprint is 26:b4:7d:98:3e:ab:19:ba:08:c9:46:9b:fb:12:5d:72.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '[10.24.24.22]:50024' (RSA) to the list of known hosts.
[email protected]'s password:
mysqlslave12.tar.gz 100% 215KB 215.2KB/s 00:00
发送到Slave2:
[root@Master1 lib]# scp -P50024 mysqlslave12.tar.gz [email protected]:/var/lib/
The authenticity of host '[10.24.24.23]:50024 ([10.24.24.23]:50024)' can't be established.
RSA key fingerprint is 26:b4:7d:98:3e:ab:19:ba:08:c9:46:9b:fb:12:5d:72.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '[10.24.24.23]:50024' (RSA) to the list of known hosts.
[email protected]'s password:
mysqlslave12.tar.gz 100% 215KB 215.2KB/s 00:00
注意:此处需要开启Master2授权root远程登录
# vim /etc/ssh/sshd_config
#PermitRootLogin no
并重启ssh连接
[root@Master1 lib]# /etc/init.d/sshd restart
数据传输到Slave1、Slave2后,依次重启Master1,Slave1、Slave2的数据库
[root@Master1 lib]# /etc/init.d/mysqld restart
Stopping mysqld: [ OK ]
Starting mysqld: [ OK ]
[root@Slave1 tools]# /etc/init.d/mysqld restart
Stopping mysqld: [ OK ]
Starting mysqld: [ OK ]
[root@Slave2 tools]# /etc/init.d/mysqld restart
Stopping mysqld: [ OK ]
Starting mysqld: [ OK ]
3.创建复制用户并授权
Master1上创建slave1、slave2的复制用户
mysql> grant replication slave on *.* to 'repl_user'@'10.24.24.22' identified by 'repl_password';
mysql> grant replication slave on *.* to 'repl_user'@'10.24.24.23' identified by 'repl_password';
刷新授权表
mysql> flush privileges;
mysql> show master status;
+------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000002 | 1296 | | |
+------------------+----------+--------------+------------------+
1 row in set (0.00 sec)
然后在slave1、slave2的数据库中将Master1设为自己的主服务器
# cd /var/lib/
# tar xf mysqlslave12.tar.gz
mysql> change master to \
-> master_host='10.24.24.20',
-> master_user='repl_user',
-> master_password='repl_password',
-> master_log_file='mysql-bin.000002',
-> master_log_pos=1296;
需要注意master_log_file和master_log_pos选项，这两个值是刚才在Master1上查询到的结果
slave1、slave2上启动从服务器，并查看slave1、slave2上的从服务器运行状态
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.24.24.20
Master_User: repl_user
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 1296
Relay_Log_File: mysql-relay-bin.000002
Relay_Log_Pos: 251
Relay_Master_Log_File: mysql-bin.000002
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: mysql.%,test.%,information_schema.%
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 1296
Relay_Log_Space: 406
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:
1 row in set (0.00 sec)
至此，Master1到slave1、slave2的MYSQL主从复制已完成。
验证数据的完整性,自己创建库或者表来进行验证数据是否同步。
以上所有设置完成后，重新启动所有节点的mysql服务。
================================================
要配置MMM，需要现在所有mysql节点创建复制帐号之外的另外两个帐号
即monitor帐号和monitor agent帐号
monitor帐号是MMM管理服务器用来对所有mysql服务器做健康检查的
monitor agent帐号用来切换只读模式和同步
所有mysql节点上创建帐号并授权：
mysql> grant replication client on *.* to 'mmm_monitor'@'10.24.24.%' identified by 'monitor_password';
mysql> grant super,replication client,process on *.* to 'mmm_agent'@'10.24.24.%' identified by 'agent_password';
刷新授权表
mysql> flush privileges;
Query OK, 0 rows affected (0.00 sec)
通过yum安装MMM后，默认设置文件目录为/etc/mysql-mmm
4个配置文件：
mmm_mon.conf 仅在MMM管理端进行配置，主要用于配置一些监控参数；
mmm_common.conf 文件需要在所有的MMM集群节点进行配置，且内容完成一样，主要用于设置读、写节点的IP及配置虚拟IP；
mmm_agent.conf 也需要在所有mysql节点进行配置，用来设置每个节点的标识。
（1）Monitor MMM集群管理端和所有mysql节点配置/etc/mysql-mmm/mmm_common.conf 文件
# cp /etc/mysql-mmm/mmm_common.conf /etc/mysql-mmm/mmm_common.conf.bak
# vim /etc/mysql-mmm/mmm_common.conf
<<--------------------------------我是华丽的代码线------------------------------>>
active_master_role writer
cluster_interface eth0
pid_path /var/run/mysql-mmm/mmm_agentd.pid
bin_path /usr/libexec/mysql-mmm/
replication_user repl_user
replication_password repl_password
agent_user mmm_agent
agent_password agent_password
ip 10.24.24.20
mode master
peer db2
ip 10.24.24.21
mode master
peer db1
ip 10.24.24.22
mode slave
ip 10.24.24.23
mode slave
hosts db1, db2
ips 10.24.24.30
mode exclusive
hosts db1, db2, db3, db4
ips 10.24.24.31, 10.24.24.32, 10.24.24.33, 10.24.24.34
mode balanced
<<--------------------------------我是华丽的代码线------------------------------>>
参数配置详解
（2）所有mysql节点配置mmm_agent.conf文件
# vim /etc/mysql-mmm/mmm_agent.conf
master1节点为：
include mmm_common.conf
this db1
master2节点为：
include mmm_common.conf
this db2
slave1节点为：
include mmm_common.conf
this db3
slave2节点为：
include mmm_common.conf
this db4
（3）MMM管理节点monitor上配置mmm_mon.conf
# vim /etc/mysql-mmm/mmm_mon.conf
<<--------------------------------我是华丽的代码线------------------------------>>
include mmm_common.conf
ip 127.0.0.1
pid_path /var/run/mysql-mmm/mmm_mond.pid
bin_path /usr/libexec/mysql-mmm
status_path /var/lib/mysql-mmm/mmm_mond.status
ping_ips 10.24.24.1, 10.24.24.20, 10.24.24.21, 10.24.24.22, 10.24.24.23
flap_duration 3600
flap_count 3
auto_set_online 0
# The kill_host_bin does not exist by default, though the monitor will
# throw a warning about it missing. See the section 5.10 "Kill Host
# Functionality" in the PDF documentation.
#
# kill_host_bin /usr/libexec/mysql-mmm/monitor/kill_host
#
monitor_user mmm_monitor
monitor_password monitor_password
debug 0
<<--------------------------------我是华丽的代码线------------------------------>>
参数配置详解
（4）配置MMM集群所有节点设置
# vim /etc/default/mysql-mmm-agent
ENABLED=1
至此，MMM集群的4个主要配置文件配置完成，将mmm_common.conf文件从MMM集群管理节点依次复制到4个mysql节点即可；
上传配置文件
# cd /etc/mysql-mmm/
# rz
# chmod 640 mmm_common.conf
MMM集群中所有配置文件的权限最好都设置为640，否则启动MMM服务的时候可能会出错。
=======================================================
8.4.5 MMM的管理
1. MMM集群服务管理
[root@Monitor ~]# /etc/init.d/mysql-mmm-monitor
Usage: /etc/init.d/mysql-mmm-monitor {start|stop|restart|condrestart|status}
[root@Master1 ~]# /etc/init.d/mysql-mmm-agent
Usage: /etc/init.d/mysql-mmm-agent {start|stop|restart|condrestart|status}
在完成MMM集群配置后，可以通过这两个脚本来启动MMM集群
在MMM集群管理端启动mysql-mmm-monitor服务
[root@Monitor ~]# /etc/init.d/mysql-mmm-monitor start
Starting MMM Monitor Daemon: [ OK ]
在每个mysql代理端依次启动agent服务
[root@Master1 ~]# /etc/init.d/mysql-mmm-agent start
Starting MMM Agent Daemon: [ OK ]
--------------------------------------------
2. MMM基本维护管理
<<--------------------------------我是华丽的代码线------------------------------>>
[root@Monitor ~]# mmm_control help
Valid commands are:
help - show this message
ping - ping monitor
show - show status
checks [
set_online
set_offline
mode - print current mode.
set_active - switch into active mode.
set_manual - switch into manual mode.
set_passive - switch into passive mode.
move_role [--force]
(Only use --force if you know what you are doing!)
set_ip
<<--------------------------------我是华丽的代码线------------------------------>>
参数详解如下：
几个常用的MMM集群维护管理的例子：
1）查看集群的运行状态
[root@Monitor mysql-mmm]# mmm_control show
db1(10.24.24.20) master/ONLINE. Roles: reader(10.24.24.33), writer(10.24.24.30)
db2(10.24.24.21) master/ONLINE. Roles: reader(10.24.24.34)
db3(10.24.24.22) slave/ONLINE. Roles: reader(10.24.24.31)
db4(10.24.24.23) slave/ONLINE. Roles: reader(10.24.24.32)
在MMM集群中，集群节点的状态有如下几种：
2）查看MMM集群目前处于什么运行模式
[root@Monitor mysql-mmm]# mmm_control mode
ACTIVE
3）查看所有MMM集群节点的运行状态
[root@Monitor mysql-mmm]# mmm_control checks all
db4 ping [last change: 2016/09/13 15:11:34] OK
db4 mysql [last change: 2016/09/13 15:11:34] OK
db4 rep_threads [last change: 2016/09/13 15:11:34] OK
db4 rep_backlog [last change: 2016/09/13 15:11:34] OK: Backlog is null
db2 ping [last change: 2016/09/13 15:11:34] OK
db2 mysql [last change: 2016/09/13 15:11:34] OK
db2 rep_threads [last change: 2016/09/13 15:11:34] OK
db2 rep_backlog [last change: 2016/09/13 15:11:34] OK: Backlog is null
db3 ping [last change: 2016/09/13 15:11:34] OK
db3 mysql [last change: 2016/09/13 15:11:34] OK
db3 rep_threads [last change: 2016/09/13 15:11:34] OK
db3 rep_backlog [last change: 2016/09/13 15:11:34] OK: Backlog is null
db1 ping [last change: 2016/09/13 15:11:34] OK
db1 mysql [last change: 2016/09/13 15:11:34] OK
db1 rep_threads [last change: 2016/09/13 15:11:34] OK
db1 rep_backlog [last change: 2016/09/13 15:11:34] OK: Backlog is null
4）单独查看某个节点的运行状态
[root@Monitor mysql-mmm]# mmm_control checks db1
db1 ping [last change: 2016/09/13 15:11:34] OK
db1 mysql [last change: 2016/09/13 15:11:34] OK
db1 rep_threads [last change: 2016/09/13 15:11:34] OK
db1 rep_backlog [last change: 2016/09/13 15:11:34] OK: Backlog is null
=========================================================
8.4.6 测试MMM实现MYSQL高可用功能
1. 读、写分离测试
首先在master1,master2,slave1,slave2上添加远程访问授权;
mysql> grant all on *.* to 'root'@'10.24.24.%' identified by 'jzh0024';
mysql> flush privileges;
mysql> select user,host from mysql.user;
通过可写的VIP登录到了Master1节点，创建一张表mmm_test,并插入一条数据；
[root@mysql01 ~]# mysql -uroot -pjzh0024 -h 10.24.24.30
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 1701
Server version: 5.1.73-log Source distribution
mysql>
mysql> show variables like "%hostname%";
+---------------+---------+
| Variable_name | Value |
+---------------+---------+
| hostname | Master1 |
+---------------+---------+
1 row in set (0.00 sec)
mysql>
mysql> create database repldb;
Query OK, 1 row affected (0.00 sec)
mysql> use repldb;
Database changed
mysql> create table mmm_test(id int,email varchar(30));
Query OK, 0 rows affected (0.01 sec)
mysql> insert into mmm_test (id,email) values(186,"[email protected]");
Query OK, 1 row affected (0.00 sec)
mysql> select * from mmm_test;
+------+----------------+
| id | email |
+------+----------------+
| 186 | [email protected] |
+------+----------------+
1 row in set (0.00 sec)
此时可以登录Master2节点、slave1节点,slave2节点，查看数据是否同步。
mysql> show databases;
mysql> use repldb;
mysql> show tables;
mysql> select * from mmm_test;
发现数据已在Master2节点、slave1节点,slave2节点上全部同步。
-------------------------------------------
接着仍在mysql远程客户端上通过MMM提供的只读VIP登录MYSQL集群，
[root@mysql01 ~]# mysql -uroot -pjzh0024 -h 10.24.24.32
mysql> show variables like "%hostname%";
+---------------+--------+
| Variable_name | Value |
+---------------+--------+
| hostname | Slave1 |
+---------------+--------+
1 row in set (0.00 sec)
mysql> use repldb;
mysql> create table mmm_test1(id int,email varchar(100));
未测试通，有问题，可以写入数据；
-------------------------------------------
2. 故障转移功能测试
先检查MMM目前的集群运行状态
[root@Monitor mysql-mmm]# mmm_control show
db1(10.24.24.20) master/ONLINE. Roles: reader(10.24.24.33), writer(10.24.24.30)
db2(10.24.24.21) master/ONLINE. Roles: reader(10.24.24.34)
db3(10.24.24.22) slave/ONLINE. Roles: reader(10.24.24.31)
db4(10.24.24.23) slave/ONLINE. Roles: reader(10.24.24.32)
关闭Master1节点的mysql服务，在查看MMM集群运行状态
http://ywliyq.blog.51cto.com/11433965/1856974