keepalived实现高可用集群（apache测试）

1.配置前准备

一台yum源服务器
两个虚拟机server4,server5
实验系统是redhat6.5
server4:172.25.254.4
server5:172.25.254.5

2.软件安装

1.配置yum源
server4和server5都要这样配置

[Server]
name=Server
baseurl=http://172.25.254.65/rhel6.5/Server
gpgcheck=0

[ScalableFileSystem]
name=ScalableFileSystem
baseurl=http://172.25.254.65/rhel6.5/ScalableFileSystem
gpgcheck=0

[HighAvailability]
name=HighAvailability
baseurl=http://172.25.254.65/rhel6.5/HighAvailability
gpgcheck=0

[LoadBalancer]
name=LoadBalancer
baseurl=http://172.25.254.65/rhel6.5/LoadBalancer
gpgcheck=0

[ResilientStorage]
name=ResilientStorage
baseurl=http://172.25.254.65/rhel6.5/ResilientStorage
gpgcheck=0

2.安装keepalived和apache,并且修改httpd的主页

yum install keepalived httpd -y
echo `hostname`>/var/www/html/index.html

3.修改keepalived的配置文件
server4: cat /etc/keepalived/keepalived.conf

global_defs {
    notification_email {  #接受邮件提醒
    root@localhost
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 127.0.0.1     #邮件服务器
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}

vrrp_instance VI_1 {
    state MASTER     #主服务
    interface eth0   #网卡端口
    virtual_router_id 51 
    priority 100     #优先级
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
    172.25.254.88/24 dev eth0 label eth0:1 #VIP
    }
}

server5： cat /etc/keepalived/keepalived.conf

global_defs {
   notification_email {
    root@172.25.254.4
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}

vrrp_instance VI_1 {
    state BACKUP   #备用服务器
    interface eth0 #网卡端口
    virtual_router_id 51 #虚拟路由值
    priority 80    #优先级
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
    172.25.254.88/24 dev eth0 label eth0:1 #VIP
    }
}

！并没有启动服务！

3.启动服务配置

至此，keepalived的配置已经完成，当主服务器(server4)的keepalived被关掉时，VIP会转至另一个结点上(server5),当主服务器被修好后，VIP有会回到主服务器上。
但是并没有实现检测httpd服务的功能，所以编写一个脚本来实现我们想要的效果。

server4: vim /etc/keepalived/bind_httpd.sh

#!/bin/bash
while true
do
curl localhost >/dev/null   #测试本机的httpd
if [ "$?" != "0" ]; then    #如果未正常运行
    /etc/init.d/httpd restart>>/var/log/httpd.log #尝试重启
    if [ "$?" != "0" ]; then #若重启失败，则关掉keepalived
        /etc/init.d/keepalived stop>/dev/null
        echo 'Error:httpd stop!'>>/var/log/httpd.log
    else            ##重启成功，则刷新keepalived
        /etc/init.d/keepalived restart >/dev/null
        echo 'Warning:httpd restart success!'>>/var/log/httpd.log
    fi
else       #如果httpd没有问题，则检测keepalived
    /etc/init.d/keepalived status>>/dev/null
    if [ "$?" != "0" ]; then  #如果keepalived被关闭，尝试重启
        /etc/init.d/keepalived restart>>/dev/null
        if [ "$?" != "0" ]; then #重启失败，日至报警
            echo 'Error:keepalived stop'>>/var/log/httpd.log
        else #重启成功，继续服务
            echo 'Warning:keepalived restart success!'>>/var/log/httpd.log
        fi
    fi
fi
sleep 5   #每5s检测一次
done

给写好的脚本添加权限并执行

chmod +x /etc/keepalived/bind_httpd.sh
echo '/etc/keepalived/bind_httpd.sh' >>/etc/rc.d/rc.local #开机便运行
reboot      #重启便执行（启动服务）

在这部分server5的操作和萨尔ver4相同。

4.测试结果

curl 172.25.254.88

>>server4

无论关调keepalived或是httpd，服务都会自起并恢复，想要看到服务器切换的效果，就把server4上的httpd卸载掉：

yum remove httpd -y

然后就发现server4上的keepalived也关掉了，VIP已经转移到server5上