ITPub博客

首页 > 数据库 > MySQL > linux上mysql MM(双主)及keepalived搭建

linux上mysql MM(双主)及keepalived搭建

原创 MySQL 作者:sqysl 时间:2020-01-29 13:21:26 0 删除 编辑

一、主备机IP及VIP规划:
master1 10.1.1.14 VIP 10.1.1.16
master2    10.1.1.15 VIP 10.1.1.16

二、mysql MM配置
1.修改master1的my.cnf
# vi /etc/my.cnf
[mysqld]
basedir = /usr/local/mysql/
datadir=/usr/local/mysql/data
log-error=/usr/local/mysql/mysqld.log
port = 3306
socket=/usr/local/mysql/mysql.sock
pid-file=/usr/local/mysql/mysql.pid

expire-logs-days=10

#binlog-do-db=db1
#binlog-ignore-db=db2

server-id = 1
log-bin = binlog
relay_log = relay-bin
log_slave_updates =1
auto_increment_increment=2
auto_increment_offset=1

2.修改master2的my.cnf
# vim /etc/my.cnf
[mysqld]
basedir = /usr/local/mysql/
datadir=/usr/local/mysql/data
log-error=/usr/local/mysql/mysqld.log
port = 3306
socket=/usr/local/mysql/mysql.sock
pid-file=/usr/local/mysql/mysql.pid

expire-logs-days=10

#binlog-do-db=db1
#binlog-ignore-db=db2

server-id=2
relay_log=relay-bin
log_bin =binlog
log_slave_updates =1
auto_increment_increment=2
auto_increment_offset=2


3.创建master1复制账号
  grant replication slave,replication client on *.* to 'repl'@'10.1.1.15' identified by 'repl';
  
4.创建master2复制账号
  grant replication slave,replication client on *.* to 'repl'@'10.1.1.14' identified by 'repl';
  
5.为master1配置master
  show master status;
  change master to master_host='10.1.1.15',master_user='repl',master_password='repl',master_log_file='binlog.000005',master_log_pos=154;


6.为master2配置master
  show master status;
  change master to master_host='10.1.1.14',master_user='repl',master_password='repl',master_log_file='binlog.000001',master_log_pos=154;
  
7.启动slave
  master1:
  start slave;
  master2:
  start slave;

三、keepalived配置
1.编辑master1的keepalived配置文件
  #vi /etc/keepalived/keepalived.conf

! Configuration File for keepalived
global_defs {
    #配置告警通知邮箱,可以配置多个
   notification_email {
    root@localhost
   }
   #配置邮件发送目标地址
   notification_email_from mysql@xiaomi.com
   #配置smtp服务器地址,其必须存在
   smtp_server 10.1.1.11
   #配置连接smtp服务器的超时时间
   smtp_connect_timeout 30
   #设置运行Keepalived实例的标识,其将显示于邮件标题中
   router_id mysql_ha
}
#监控脚本
vrrp_script chk_mysql {
    script "/etc/keepalived/check_mysql.sh"
    interval 2
    weight 2
}
#配置VRRP实例,实例命名任意
vrrp_instance mysql-ha {
    #配置Keepalived角色,MASTER为主机 BACKUP为备机,此处两个都设置为BACKUP
    state BACKUP 
    #配置keepalived监测的网络接口
    interface eth0
    #虚拟路由标识,其为一个(1-255)的数字,一个VRRP实例中主机的该ID必须相同
    virtual_router_id 66
    #服务器优先级,数字越大优先级越高,一个实例中主服务器优先级要高于备服务器
    priority 50  
    #配置主备服务器间同步检查的时间间隔(秒)
    advert_int 1
    #配置服务器抢占模式,这里配置为非抢占模式(只需对master1配置即可)
    nopreempt
    #配置验证类型和密码
    authentication {
        #两种验证类型{PASS|HA}
        auth_type PASS
        #指定验证密码,一个实例中的主备服务器密码要一样
        auth_pass centos
    }
    track_script {
    #指定执行监控的服务
        chk_mysql
    }
    #配置虚拟IP,可指定有多个,每个占一行
    virtual_ipaddress {
    10.1.1.16
    }
}

2.编辑master1心跳检测脚本:
#vi /etc/keepalived/check_mysql.sh 

#!/bin/bash
#This scripts is check for Mysql Slave status
counter=$(netstat -na|grep "LISTEN"|grep "3311"|wc -l)
if [ "${counter}" -eq 0 ]; then
    service keepalived stop
    killall keepalived
fi
ping 10.1.1.14 -w1 -c1 &>/dev/null
if [ $? -ne 0 ]
then
    systemctl stop keepalived
    killall keepalived
fi

3.编辑master2的keepalived配置文件
# vim /etc/keepalived/keepalived.conf 

! Configuration File for keepalived
global_defs {
   #配置告警通知邮箱,可以配置多个
   notification_email {
    root@localhost
   }
   #配置邮件发送目标地址
   notification_email_from mysql@xiaomi.com
   #配置smtp服务器地址,其必须存在
   smtp_server 10.1.1.11
   #配置连接smtp服务器的超时时间
   smtp_connect_timeout 30
   #设置运行Keepalived实例的标识,其将显示于邮件标题中
   router_id mysql_ha
}
# 监控监本
vrrp_script chk_mysql {
    script "/etc/keepalived/check_mysql.sh"
    interval 2
    weight 2
}
#配置VRRP实例,实例命名任意
vrrp_instance mysql-ha {
    #配置Keepalived角色,MASTER为主机 BACKUP为备机,此处两个都设置为BACKUP
    state BACKUP
    #配置keepalived监测的网络接口
    interface eth0
    #虚拟路由标识,其为一个(1-255)的数字,一个VRRP实例中主机的该ID必须相同
    virtual_router_id 66
    #服务器优先级,数字越大优先级越高,一个实例中主服务器优先级要高于备服务器
    priority 49
    #配置主备服务器间同步检查的时间间隔(秒)
    advert_int 1
    #配置服务器抢占模式,这里配置为非抢占模式(只需对master1配置即可)
    #nopreempt
    #配置验证类型和密码
    authentication {
        #两种验证类型{PASS|HA}
        auth_type PASS
        #指定验证密码,一个实例中的主备服务器密码要一样
        auth_pass centos
    }
    track_script {
    #指定执行监控的服务
        chk_mysql  
    }
    #配置虚拟IP,可指定有多个,每个占一行
    virtual_ipaddress {
    10.1.1.16
    }
}

4.编辑master2检测脚本
# vim /etc/keepalived/check_mysql.sh 

#!/bin/bash
#This scripts is check for Mysql Slave status
counter=$(netstat -na|grep "LISTEN"|grep "3306"|wc -l)
if [ "${counter}" -eq 0 ]; then
    systemctl stop keepalived
    killall keepalived
fi
ping 10.1.1.15 -w1 -c1 &>/dev/null
if [ $? -ne 0 ]
then
    service keepalived stop
    killall keepalived
fi

5.vip漂移检测
1)master1和master2上同时开启keepalived和mysql
  #service keepalived start
  #service mysqld start
2)查看master1上ip地址
  ip addr
3)登录10.1.1.16上的mysql
  mysql -uusername -ppassword -h10.1.1.16 -P3311
4)停掉master1上的mysql服务
  service mysqld stop
5)观察master1和master2上的ip地址
  ip addr
6)继续在3)中的session中运行mysql命令,看看发生了什么
  mysql> use information_schema;
 


来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/8484829/viewspace-2674315/,如需转载,请注明出处,否则将追究法律责任。

全部评论
上世纪90年代初至今,一直专注于数据库应用及研究,擅长数据库优化及设计等。http://blog.itpub.net/8484829;https://blog.csdn.net/tuning_optmization;https://www.cnblogs.com/lhdz_bj/

注册时间:2008-06-27

  • 博文量
    356
  • 访问量
    569581