atlas代理mysql_Mysql中间件代理 Atlas

本文接上文介绍atlas的安装配置，以及借助keepalive软件实现altas的ha，避免由于atlas故障，导致数据库服务中断的情况出现：

一：环境介绍

Atlas 主：192.168.1.12/24

Atlas 从：192.168.1.81/24

Atlas vip:  192.168.1.230/24

Atlas write:

master：192.168.1.225/24

Slave1：192.168.1.226/24

Slave2：192.168.1.227/24

管理节点：192.168.1.12/24

Mha vip:  192.168.1.231/24

Atlas  read:

Slave2：192.168.1.227/24

Slave3：192.168.1.228/24

二：安装atlas，12服务器及81服务器执行同样的操作

# yum -y install openssl openssl-devel

# wget https://github.com/Qihoo360/Atlas/releases/download/2.1/Atlas-2.1.el6.x86_64.rpm

# rpm -ivh Atlas-2.1.el6.x86_64.rpm

# cd /usr/local/mysql-proxy/conf/

# /usr/local/mysql-proxy/bin/encrypt 123456

/iZxz+0GRoA=

# grep -v '^#' /usr/local/mysql-proxy/conf/test.cnf  |grep -v '^$'

[mysql-proxy]

admin-username = admin

admin-password = admin

proxy-backend-addresses = 192.168.1.231:3306

proxy-read-only-backend-addresses = 192.168.1.227:3306,192.168.1.228:3306

pwds = root:/iZxz+0GRoA=

daemon = true

keepalive = true

event-threads = 8

log-level = message

log-path = /usr/local/mysql-proxy/log

proxy-address = 0.0.0.0:1234

admin-address = 0.0.0.0:2345

# /usr/local/mysql-proxy/bin/mysql-proxyd test start

OK: MySQL-Proxy of test is started

# netstat -ntpl |grep mysql-proxy

tcp        0      0 0.0.0.0:2345                0.0.0.0:*                   LISTEN      29643/mysql-proxy

tcp        0      0 0.0.0.0:1234                0.0.0.0:*                   LISTEN      29643/mysql-proxy

# tail -f /usr/local/mysql-proxy/log/test.log

2014-10-29 11:12:11: (message) mysql-proxy 0.8.2 started - instance: test

2014-10-29 11:12:11: (message) proxy listening on port 0.0.0.0:1234

2014-10-29 11:12:11: (message) added read/write backend: 192.168.1.231:3306

2014-10-29 11:12:11: (message) added read-only backend: 192.168.1.227:3306

2014-10-29 11:12:11: (message) chassis-unix-daemon.c:138: [angel] we try to keep PID=29643 alive2014-10-29 11:12:11: (message) added read-only backend: 192.168.1.228:3306

2014-10-29 11:12:11: (message) chassis-event-thread.c:235: starting 8 threads

# /usr/local/mysql/bin/mysql -h 192.168.1.12 -P 1234 -u root -p123456

# /usr/local/mysql/bin/mysql -h 192.168.1.12-P 2345 -uadmin -padmin

三：安装keepalived

此处以12服务器为例，81服务器执行同样的操作

# cd /usr/local/src/

# wget http://www.keepalived.org/software/keepalived-1.2.13.tar.gz

# tar -zxvpf keepalived-1.2.13.tar.gz

# cd keepalived-1.2.13

# ./configure --prefix=/usr/local/keepalived  && make && make install

四：配置12服务器keepalive

1：keepalive主配置文件

# cat /usr/local/keepalived/etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {

router_id  LVS_DEVEL

}

vrrp_script Monitor_Atlas {

script "/usr/local/scripts/monitor_atlas.sh"

interval 2

weight 2

}

vrrp_instance VI_1{

state MASTER

interface eth0

virtual_router_id 51

mcast_src_ip 192.168.1.12

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass password_123

}

track_script {

Monitor_Atlas

}

virtual_ipaddress {

192.168.1.230

}

}

2：监控atlas脚本

# cat /usr/local/scripts/monitor_atlas.sh

#!/bin/bash

if  [ $(ps -ef |grep 'mysql-proxy' |grep -v 'grep' |wc -l) -eq "0" ];then

/usr/local/mysql-proxy/bin/mysql-proxyd test start

sleep 5

if [ $(ps -ef |grep 'mysql-proxy' |grep -v 'grep' |wc -l) -eq "0"  ];then

killall -9 keepalived

service network restart

fi

fi

# chmod +x /usr/local/scripts/monitor_atlas.sh

五：配置81服务器keepalive

1：keepalive主配置文件

# cat /usr/local/keepalived/etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {

router_id  LVS_DEVEL

}

vrrp_script Monitor_Atlas {

script "/usr/local/scripts/monitor_atlas.sh"

interval 2

weight 2

}

vrrp_instance VI_1{

state BACKUP

interface eth0

virtual_router_id 51

mcast_src_ip 192.168.1.81

priority 90

advert_int 1

authentication {

auth_type PASS

auth_pass password_123

}

track_script {

Monitor_Atlas

}

virtual_ipaddress {

192.168.1.230

}

}

2：监控atlas脚本

# cat /usr/local/scripts/monitor_atlas.sh

#!/bin/bash

if  [ $(ps -ef |grep 'mysql-proxy' |grep -v 'grep' |wc -l) -eq "0" ];then

/usr/local/mysql-proxy/bin/mysql-proxyd test start

sleep 5

if [ $(ps -ef |grep 'mysql-proxy' |grep -v 'grep' |wc -l) -eq "0"  ];then

killall -9 keepalived

service network restart

fi

fi

# chmod +x /usr/local/scripts/monitor_atlas.sh

六：启动keepalive服务及测试

1:12服务器

# /usr/local/keepalived/sbin/keepalived -D -f /usr/local/keepalived/etc/keepalived/keepalived.conf

# tail -f /var/log/messages

2: 81服务器

# /usr/local/keepalived/sbin/keepalived -D -f /usr/local/keepalived/etc/keepalived/keepalived.conf

# tail -f /var/log/messages

3：连接atlas测试

# /usr/local/mysql/bin/mysql -h 192.168.1.230 -P 1234 -uroot -p123456

mysql> select user,host,password from mysql.user;

# /usr/local/mysql/bin/mysql -h 192.168.1.230 -P 2345 -uadmin -padmin

mysql>  SELECT * FROM backends;

4：故障测试

12服务器上关闭keepalive服务

81服务器的keepalive会自动接管vip服务

# tail -f /var/log/messages