RedHat 7配置keepalived实现HAproxy群集

在之前的博文中，我搭建了一个1台HAproxy+2台Apache的WEB负载均衡测试环境，然而1台HAproxy终归存在单点故障的风险，如果能使用2台HAproxy做高可用，形成主备群集，对外提供一个VIP在两台HAproxy之间浮动，一旦主的故障，备机立马接管VIP，保持HAproxy仍旧可用，那就比较完美了，其实很简单，Keepalived就能实现这样的功能，且在RedHat 7的安装光盘中自带了keepalived的RPM包，通过yum安装非常方便。通过在两台HAproxy上分别安装Keepalived，形成一主一备，使用VRRP协议检测心跳信号，VIP可以在两台HAproxy之间浮动，实现HAproxy的高可用群集架构。架构示意图如下：

关于Apache和HAproxy的安装配置可以参考本人之前的博文：

《RedHat 7配置HAProxy实现Web负载均衡》

http://jiangjianlong.blog.51cto.com/3735273/1887279

《RedHat 7 安装配置Apache 2.4》

http://jiangjianlong.blog.51cto.com/3735273/1858650

一、测试环境

计算机名	操作系统	IP地址	VIP	安装的软件
RH7-HAproxy01	Redhat 7.2	192.168.10.19	192.168.10.18	Haproxy1.5.14、keepalived 1.2.13
RH7-HAproxy02	Redhat 7.2	192.168.10.20	192.168.10.18	Haproxy1.5.14、keepalived 1.2.13
RH7-Apache01	Redhat 7.2	192.168.10.21	——	Apache httpd 2.4
RH7-Apache02	Redhat 7.2	192.168.10.22	——	Apache httpd 2.4

二、安装配置keepalived

1、在两台HAproxy上分别使用yum install keepalived命令安装keepalived

[[email protected] ~]# yum install keepalived

2、编辑RH7-HAproxy01的keepalived配置文件，将其配置成MASTER

[[email protected]~]# cd /etc/keepalived/

[[email protected]]# ls

keepalived.conf

[[email protected]]# cp keepalived.conf keepalived.conf.BAK

[[email protected]]# ls

keepalived.conf keepalived.conf.BAK

[[email protected]]# vi keepalived.conf

! Configuration Filefor keepalived
 
global_defs {
   notification_email {
     [email protected]
     [email protected]
     [email protected]
   }
   [email protected]
   smtp_server 192.168.200.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
 
#监测haproxy进程状态，每2秒执行一次
vrrp_scriptchk_haproxy {
    script"/etc/keepalived/chk_haproxy.sh"
    interval 2
    weight 2
    fall 3
}
 
vrrp_instance VI_1 {
    state MASTER  #将HR7-HAproxy01标示为MASTER
    interface eno50332208 #指定网络接口
    virtual_router_id 51
    priority 100  #MASTER权重要高于BACKUP
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.10.18   #VIP
    }
    track_script {
        chk_haproxy #监测haproxy进程状态
}
}

3、编辑RH7-HAproxy02的keepalived配置文件，将其配置成BACKUP

[[email protected] ~]# cd /etc/keepalived/

[[email protected] keepalived]# ls

keepalived.conf

[[email protected] keepalived]# cp keepalived.confkeepalived.conf.BAK

[[email protected] keepalived]# ls

keepalived.conf keepalived.conf.BAK

[[email protected]]# vi keepalived.conf

[[email protected]~]# vi /etc/keepalived/keepalived.conf

global_defs {
   notification_email {
     [email protected]
     [email protected]
     [email protected]
   }
   [email protected]
   smtp_server 192.168.200.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
 
#监测haproxy进程状态，每2秒执行一次
vrrp_scriptchk_haproxy {
    script"/etc/keepalived/chk_haproxy.sh"
    interval 2
    weight 2
}
 
 
vrrp_instance VI_1 {
    state BACKUP  #将HR7-HAproxy02标识为BACKUP
    interface eno50332208 #指定网络接口
    virtual_router_id 51
    priority 99  #MASTER权重要高于BACKUP
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
 
    virtual_ipaddress {
        192.168.10.18   #VIP
    }
    track_script {
        chk_haproxy #监测haproxy进程状态
}
}

4、在2台HAproxy上创建chk_haproxy.sh脚本，内容如下：

vi /etc/keepalived/chk_haproxy.sh

#!/bin/bash
status=$(ps aux|grephaproxy | grep -v grep | grep -v bash | wc -l)
if [ "${status}" = "0" ]; then  #判断haproxy是否已经启动
   systemctl start haproxy #如果没有启动，则启动haproxy程序
 
sleep 3  #睡眠3秒钟以等待haproxy完全启动
 
status2=$(psaux|grep haproxy | grep -v grep | grep -v bash |wc -l)
    if [ "${status2}" ="0"  ]; then
    systemctl stop keepalived
   #如果haproxy还是没有启动，则将keepalived停掉，这样VIP会自动漂移到另外一台haproxy
    fi
fi