LVS+Keepalived

说明：本文档仅围绕lvs+keepalived如何实现负载均衡、故障剔除、后端realserver健康监测、主备切换邮件通知;而防火墙、网络(路由交换)、后端数据存储、内外网暂未考虑; 一、环境准备： 1.操作系统

CentOS6.4-x86_64

2.软件版本：

ipvsadm-1.25-10.el6.x86_64
keepalived-1.2.7-3.el6.x86_64
httpd-2.2.15-26.el6.centos.x86_64

3.实验拓扑： [

1](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/1.png)

4.时间同步：

node1:
\[root@node1 ~\]# ntpdate 203.117.180.36
\[root@node1 ~\]# echo "*/10 * * * * /usr/sbin/ntpdate 203.117.180.36" >> /etc/crontab
node2:
\[root@node2 ~\]# ntpdate 203.117.180.36
\[root@node1 ~\]# echo "*/10 * * * * /usr/sbin/ntpdate 203.117.180.36" >> /etc/crontab
master:
\[root@master ~\]# ntpdate 203.117.180.36
\[root@master ~\]# echo "*/10 * * * * /usr/sbin/ntpdate 203.117.180.36" >> /etc/crontab
Slave:
\[root@slave ~\]# ntpdate 203.117.180.36
\[root@slave ~\]# echo "*/10 * * * * /usr/sbin/ntpdate 203.117.180.36" >> /etc/crontab

5.主机名相互解析：

node1:
\[root@node1 ~\]# cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4 
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6 
192.168.254.201    node1.test.com    node1 
192.168.254.202    node2.test.com    node2
node2:
\[root@node2 ~\]# cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4 
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6 
192.168.254.201    node1.test.com    node1 
192.168.254.202    node2.test.com    node2

6.安装yum源：(其他三台主机上面同样执行以下两条命令即可，前提:能上网)

node1:
\[root@node1~\]#
rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
\[root@node1 ~\]# rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

二、**web**节点安装配置：安装web服务并执行realserver.sh,为lo:0绑定VIP地址192.168.254.200，抑制ARP广播; node1/node2配置：

\[root@node1 ~\]# yum install -y httpd        
\[root@node1 ~\]# echo "<h1>Rs1.test.com</h1>" > /var/www/html/index.html
\[root@node1 ~\]# service httpd start
\[root@node1 ~\]# chkconfig httpd on
\[root@node1 ~\]# mkdir script
\[root@node1 ~\]# cd script/
\[root@node1 ~\]# vim script/realserver.sh
#->LVS客户端配置脚本realserver.sh：
#!/bin/bash 
\# Script to start LVS DR real server.  
\# description: LVS DR real server  

.  /etc/rc.d/init.d/functions
VIP=192.168.254.200 
host=`/bin/hostname`
case "$1" in 
start)  
       # Start LVS-DR real server on this machine.  
       /sbin/ifconfig lo down  
        /sbin/ifconfig lo up  
        echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore  
        echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce  
        echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore  
        echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce
        /sbin/ifconfig lo:0 $VIP broadcast $VIP netmask 255.255.255.255 up 
        /sbin/route add -host $VIP dev lo:0
;; 

stop)

        # Stop LVS-DR real server loopback device(s). 
        /sbin/ifconfig lo:0 down  
        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore  
        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce  
        echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore  
        echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce
;; 

status)
        # Status of LVS-DR real server. 
        islothere=`/sbin/ifconfig lo:0 | grep $VIP`  
        isrothere=\`netstat -rn | grep "lo:0" | grep $VIP\`  
        if \[ ! "$islothere" -o ! "isrothere" \];then  
            # Either the route or the lo:0 device  
            # not found.  
            echo "LVS-DR real server Stopped."  
        else  
           echo "LVS-DR real server Running."  
        fi  
;;  
*)  
            # Invalid entry.  
            echo "$0: Usage: $0 {start|status|stop}"  
            exit 1  

;;  
esac
\[root@node1 ~\]# chmod +x script/realserver.sh
\[root@node1 ~\]# ./script/realserver.sh start

#->将该脚本scp到node2节点执行./script/realserver.sh start即可

#->如果服务器重启，那还需要手动的去执行这个脚本，服务便不可使用了，所以可以将realserver.sh加入到开机启动项中;

\[root@node1 ~\]# vim /etc/rc.local
/bin/bash /root/script/realserver.sh start
\[root@node1 ~\]# scp /script/realserver 192.168.254.46:/root/script
#->效果如下：
[![2](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/2.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/2.png) [![3](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/3.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/3.png)
#->客户端访问测试：
[![4](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/4.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/4.png)

三、**LVS-DR-Master/Slave**安装配置： 3.1.master/slave**都安装keepalived和**ipvsadm :

master:
\[root@master ~\]# yum install -y ipvsadm keepalived
\[root@master ~\]# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak  #->修改配置文件之前最好将该配置文件备份。避免后续出现问题
Slave:
\[root@slave ~\]# yum install -y ipvsadm keepalived
\[root@slave ~\]# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak

**3.2.****在****Lvs-DR-Master****上面的配置：**

\[root@master ~\]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   notification_email {
    2399447849@qq.com               #->#设置报警邮件地址，可以设置多个，每行一个
   }
   notification\_email\_from root@localhost.localdomain
   smtp_server 127.0.0.1            #->设置smtp server的地址
   smtp\_connect\_timeout 30          #->设置连接smtp server的超时时间
   router\_id LVS\_DEVEL              #->表示运行keepalived服务器的一个标识。发邮件时显示在邮件主题的信息
}

vrrp\_instance VI\_1 {
    state MASTER           #->指定keepalived的角色，MASTER表示此主机是主服务器，BACKUP表示此主机是备用服务器
    interface eth0         #->指定HA监测网络的接口
    virtual\_router\_id 60  #->虚拟路由标识，这个标识是一个数字,同一个vrrp实例使用唯一的标识。即同一vrrp_instance下,MASTER和BACKUP>必须是一致的
    priority 101          #->定义优先级，数字越大，优先级越高，在同一个vrrp_instance下，MASTER的优先级必须大于BACKUP的优先级
    advert_int 1          #->设定MASTER与BACKUP负载均衡器之间同步检查的时间间隔，单位是秒
    authentication {      #->设置验证类型和密码
        auth_type PASS    #->设置验证类型，主要有PASS和AH两种
        auth\_pass 1111    #->设置验证密码，在同一个vrrp\_instance下，MASTER与BACKUP必须使用相同的密码才能正常通信
   }

    virtual_ipaddress {   #->设置虚拟IP地址，可以设置多个虚拟IP地址，每行一个
        192.168.254.200   #->客户端通过访问的就是该IP地址
    }
}

virtual_server 192.168.254.200 80 {  #->设置虚拟服务器，需要指定虚拟IP地址和服务端口，IP与端口之间用空格隔开
    delay_loop 6                       #->设置运行情况检查时间，单位是秒
    lb_algo rr                          #->设置负载调度算法，这里设置为rr，即轮询算法
    lb_kind DR                          #->设置LVS实现负载均衡的机制，有NAT、TUN、DR三个模式可选
nat_mask 255.255.255.0
#persistence_timeout 50      #->会话保持时间，单位是秒。这个选项对动态网页是非常有用的，为集群系统中的session共享提供了一个很好>的解决方案。
                               #->有了这个会话保持功能，用户的请求会被一直分发到某个服务节点，直到超过这个会话的保持时间。
                               #->需要注意的是，这个会话保持时间是最大无响应超时时间，也就是说，用户在操作动态页面时，如果50秒内没
有执行任何操作，
    protocol TCP                  #->指定转发协议类型，有TCP和UDP两种
    real_server 192.168.254.45 80  {     #->配置服务节点1，需要指定real server的真实IP地址和端口，IP与端口之间用空格隔开
        weight 1                           #->配置服务节点的权值，权值大小用数字表示，数字越大，权值越高，设置权值大小可以为不同性能的服务器
                                            #->分配不同的负载，可以为性能高的服务器设置较高的权值，而为性能较低的服务器设置相对较低的权值，这样
才能合理地利用和分配系统资源
        HTTP_GET {                         #->realserver的状态检测设置部分，单位是秒
            url {                         
              path /
               status_code 200            #->状态码定义
            }
            connect_timeout 3             #->表示3秒无响应超时
            nb\_get\_retry 3                #->表示重试次数
            delay\_before\_retry 3         #->表示重试间隔
                  }

    }

    real_server 192.168.254.46 80 {
        weight 1
        HTTP_GET {
            url {
              path /
               status_code 200
           }
            connect_timeout 3
            nb\_get\_retry 3
            delay\_before\_retry 3
        }
    }
}

#->根据实验的拓扑可知后端的realserver有两台，所以上面定义了两个realserver的配置

3.3**在Lvs-DR-Slave上面的配置：**

[root@master ~]# scp /etc/keepalived/keepalived.conf 192.168.254.48:/etc/keepalived/ #->将master上面的配置复制至slave,然后稍作修改：
[root@slave ~]# vim /etc/keepalived/keepalived.conf

#->配置文件各个参数上面以解释，这里只对需要修改的作说明：

! Configuration File for keepalived


global_defs {
   notification_email {
    2399447849@qq.com
   }
   notification\_email\_from root
   smtp_server 127.0.0.1
   smtp\_connect\_timeout 30
   router\_id LVS\_DEVEL

}

 

vrrp\_instance VI\_1 {
    state BACKUP               #->指定该服务器的keepalived角色为BACKUP(备用服务器)
    interface eth0
    virtual\_router\_id 60     
    priority 100               #->在同一个vrrp_instance下，MASTER的优先级必须大于BACKUP的优先级
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }

    virtual_ipaddress {
        192.168.254.200
    }
}

 

virtual_server 192.168.254.200 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
   nat_mask 255.255.255.0
    #persistence_timeout 50
protocol TCP
   real_server 192.168.254.45 80  {
        weight 1
        HTTP_GET {
            url {
              path /
               status_code 200
            }
            connect_timeout 3
            nb\_get\_retry 3
            delay\_before\_retry 3
                 }
    }
    real_server 192.168.254.46 80 {
        weight 1
        HTTP_GET {
            url {
             path /
              status_code 200
            }
            connect_timeout 3
            nb\_get\_retry 3
            delay\_before\_retry 3
        }
    }
}

3.4.**启动keepalived (启动过程中观察日志**)

master:
\[root@master ~\]# service keepalived start && tail -f /var/log/messages
...........
...........
Oct 29 21:42:14 master Keepalived_healthcheckers\[31358\]: Opening file '/etc/keepalived/keepalived.conf'.
Oct 29 21:42:14 master Keepalived_healthcheckers\[31358\]: Configuration is using : 16384 Bytes
Oct 29 21:42:14 master Keepalived_healthcheckers\[31358\]: Using LinkWatch kernel netlink reflector...
Oct 29 21:42:14 master Keepalived_healthcheckers\[31358\]: Activating healthchecker for service \[192.168.254.45\]:80
Oct 29 21:42:14 master Keepalived_healthcheckers\[31358\]: Activating healthchecker for service \[192.168.254.46\]:80
Oct 29 21:42:15 master Keepalived\_vrrp\[31359\]: VRRP\_Instance(VI_1) Transition to MASTER STATE
Oct 29 21:42:16 master Keepalived\_vrrp\[31359\]: VRRP\_Instance(VI_1) Entering MASTER STATE  #->主服务器状态
Oct 29 21:42:16 master Keepalived\_vrrp\[31359\]: VRRP\_Instance(VI_1) setting protocol VIPs.
Oct 29 21:42:16 master Keepalived\_vrrp\[31359\]: VRRP\_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.254.200
Oct 29 21:42:16 master Keepalived_healthcheckers\[31358\]: Netlink reflector reports IP 192.168.254.200 added
Oct 29 21:42:21 master Keepalived\_vrrp\[31359\]: VRRP\_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.254.200
...........
...........
\[root@master ~\]# ipvsadm -L –n       #->LVS状态
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -\> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.254.200:80 rr
  -\> 192.168.254.45:80            Route   1      0          0        
  -\> 192.168.254.46:80            Route   1      0          0        
\[root@master ~\]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid\_lft forever preferred\_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER\_UP> mtu 1500 qdisc pfifo\_fast state UP qlen 1000
   link/ether 00:0c:29:5d:7d:94 brd ff:ff:ff:ff:ff:ff
    inet 192.168.254.47/24 brd 192.168.254.255 scope global eth0
    inet 192.168.254.200/32 scope global eth0    #->此时VIP在master上面
    inet6 fe80::20c:29ff:fe5d:7d94/64 scope link
       valid\_lft forever preferred\_lft forever
\[root@master ~\]#
Slave:
\[root@slave ~\]# service keepalived start && tail -f /var/log/messages
...........
...........
Oct 29 21:42:34 slave Keepalived_vrrp\[31389\]: Opening file '/etc/keepalived/keepalived.conf'.
Oct 29 21:42:34 slave Keepalived_vrrp\[31389\]: Configuration is using : 62845 Bytes
Oct 29 21:42:34 slave Keepalived_vrrp\[31389\]: Using LinkWatch kernel netlink reflector...
Oct 29 21:42:34 slave Keepalived\_vrrp\[31389\]: VRRP\_Instance(VI_1) Entering BACKUP STATE  #->备用服务器状态
Oct 29 21:42:34 slave Keepalived_vrrp\[31389\]: VRRP sockpool: \[ifindex(2), proto(112), fd(10,11)\]
Oct 29 21:42:34 slave Keepalived_healthcheckers\[31388\]: Opening file '/etc/keepalived/keepalived.conf'.
Oct 29 21:42:34 slave Keepalived_healthcheckers\[31388\]: Configuration is using : 16384 Bytes
Oct 29 21:42:34 slave Keepalived_healthcheckers\[31388\]: Using LinkWatch kernel netlink reflector...
Oct 29 21:42:34 slave Keepalived_healthcheckers\[31388\]: Activating healthchecker for service \[192.168.254.45\]:80
Oct 29 21:42:34 slave Keepalived_healthcheckers\[31388\]: Activating healthchecker for service \[192.168.254.46\]:80
...........
...........

**3.5.****测试：** [![5](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/5.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/5.png) **3.6.****模拟故障**： (1)停掉node1节点的web服务：

\[root@node1 ~\]# service httpd stop
停止 httpd：                                               \[确定\]
\[root@node1 ~\]#

(2)查看一下报警邮件： [![6](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/6.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/6.png) (3)再在前端调度器上查看一下LVS状态： [![7](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/7.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/7.png) 很明显那台出现问题的realserver条目已经被剔除了 (4)恢复node1节点上的web服务：

\[root@node1 ~\]# service httpd start
启动 httpd：                                               \[确定\]
\[root@node1 ~\]#

(5) 查看一下报警邮件： [![8](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/8.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/8.png) (6)关闭master上面的keepalived：

\[root@master ~\]# service keepalived stop
停止 keepalived：                                          \[确定\]
\[root@master ~\]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -\> RemoteAddress:Port           Forward Weight ActiveConn InActConn
\[root@master ~\]#

(7)查看slave状态：

\[root@slave ~\]# ip a
.......
.......
    inet 192.168.254.200/32 scope global eth0     #->可见VIP已经转移到了slave上面；并且通过客户端访问仍然正常！
.......
.......
 \[root@slave ~\]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)

Prot LocalAddress:Port Scheduler Flags

  -\> RemoteAddress:Port           Forward Weight ActiveConn InActConn

TCP  192.168.254.200:80 rr
  -\> 192.168.254.45:80            Route   1      0          0        
  -\> 192.168.254.46:80            Route   1      0          0        
\[root@slave ~\]#

**通过上面的演示现在的****LVS****的高可用即前端负载均衡调度器的高可用，同时实现了对后端****realserver****监控，也实现了后端****realserver****宕机时会给管理员发送邮件；但是目前还面临几个问题：**

1.  **如果所有的****realserver****都宕机，如何处理，用户打不开就等它打不开，还是友善的提示一下？**
2.  **怎么完成维护模式****keepalived****切换？**
3.  **如何在****keepalived****主备切换时向管理员发送邮件？**

**四、****LVS+Keepalived****后续延伸：** **4.1.****所有****realserver****都宕机如何处理？** 在集群中如果所有real server全部宕机了，客户端访问时就会出现错误页面，这样是很不友好的，我们得提供一个维护页面来提醒用户，服务器正在维护，什么时间可以访问等，下面就来解决一下这个问题。解决方案有两种，一种是提供一台备用的real server当所有的服务器宕机时，提供维护页面，但这样做有点浪费服务器。另一种就是在负载均衡器上提供维护页面，这样是比较靠谱的，也比较常用。下面就来具体操作一下。 (1)在master和slave上面安装httpd

\[root@master ~\]# yum install -y httpd
\[root@slave ~\]# yum install -y httpd

(2)提供维护页面文件

\[root@master ~\]# echo "Oops ... you visit the page does not exist, the server may be maintained?" > /var/www/html/index.html
\[root@master ~\]# service httpd start
启动 httpd：                                               \[确定\]
\[root@slave ~\]# echo "Oops ... you visit the page does not exist, the server may be maintained?" > /var/www/html/index.html
\[root@slave ~\]# service httpd start
启动 httpd：                                               \[确定\]

(3)测试： [![9](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/9.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/9.png) (4)修改master/slave的keepalived配置文件：

\[root@master ~\]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
notification_email {
2399447849@qq.com
}

notification\_email\_from root
smtp_server 127.0.0.1
smtp\_connect\_timeout 30
router\_id LVS\_DEVEL
}

vrrp\_instance VI\_1 {
state MASTER
interface eth0
       virtual\_router\_id 60
       priority 101
       advert_int 1
      authentication {
       auth_type PASS
      auth_pass 1111
}

virtual_ipaddress {
     192.168.254.200
    }
}


virtual_server 192.168.254.200 80 {
delay_loop 6
lb_algo rr
lb_kind DR
nat_mask 255.255.255.0
#persistence_timeout 50
protocol TCP

real_server 192.168.254.45 80&nbsp; {
weight 1
HTTP_GET {

  url {
   path /
   status_code 200
}

connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
   }
}

real_server 192.168.254.46 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
}
}

sorry_server 127.0.0.1&nbsp;&nbsp; #->增加该配置参数，slave上面也需要添加，此处略。
}

(5)关闭所有的realserver web服务,重新启动master/slave 的keepalived:  

node1:
\[root@node1 ~\]# service httpd stop
停止 httpd：                                               \[确定\]
\[root@node1 ~\]#
node2:
\[root@node2 ~\]# service httpd stop
停止 httpd：                                               \[确定\]
\[root@node2 ~\]#
master:
\[root@master ~\]# service keepalived restart
停止 keepalived：                                          \[确定\]
正在启动 keepalived：                                      \[确定\]
\[root@master ~\]#
slave:
\[root@slave ~\]# service keepalived restart
停止 keepalived：                                          \[确定\]
正在启动 keepalived：                                      \[确定\]
\[root@slave ~\]#

(6)查看一下LVS状态： [![10](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/10.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/10.png) (7)访问测试： [![11](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/11.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/11.png) **4.2.****如何完成维护模式****keepalived** **切换？** 一般我们在测试主从切换的过程当中要么是手动停止keepalived服务，要么是手动关闭网卡，那还有其他方法实现维护模式的切换，这就是vrrp_script功能； (1)master/slave配置：**(****注****:****这里演示主服务器的配置，添加上去的在****slave****上面也需要添加以红色标注内容****)**

\[root@master ~\]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {

notification_email {
2399447849@qq.com
}

notification\_email\_from root
smtp_server 127.0.0.1
smtp\_connect\_timeout 30
router\_id LVS\_DEVEL
}

vrrp\_script chk\_schedown {       #->定义vrrp执行脚本
   script "\[ -e /etc/keepalived/down \] && exit 1 || exit 0"  #->查看是否有down文件，有就进入维护模式
   interval 1         #->监控间隔时间
   weight -5          #->降低优先级,即priority参数
   fall 2             #->失败次数
   rise 1             #->成功次数
}

vrrp\_instance VI\_1 {
state MASTER
interface eth0
virtual\_router\_id 60
priority 101
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}

virtual_ipaddress {
192.168.254.200
}

track_script {             #->脚本追踪
    chk_schedown           #->上面自定义的vrrp脚本名称
    }
}


virtual_server 192.168.254.200 80 {
delay_loop 6
lb_algo rr
lb_kind DR
nat_mask 255.255.255.0
#persistence_timeout 50
protocol TCP
 
real_server 192.168.254.45 80  {
weight 1
HTTP_GET {
url {
path /
status_code 200
}

connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
}

}

real_server 192.168.254.46 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}

connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
}
}

sorry_server 127.0.0.1
}

(2)测试： 

master:
\[root@master keepalived\]# touch down  #->新建一个down文件，进入维护模式
\[root@master keepalived\]# ll
总用量 4
-rw-r--r--. 1 root root    0 10月 30 00:16 down
-rw-r--r--. 1 root root 1513 10月 30 00:08 keepalived.conf
\[root@master keepalived\]# tail -f /var/log/messages
.......
.......
Oct 30 00:16:43 node3 Keepalived\_vrrp\[31993\]: VRRP\_Script(chk_schedown) failed
Oct 30 00:16:44 node3 Keepalived\_vrrp\[31993\]: VRRP\_Instance(VI_1) Received higher prio advert
Oct 30 00:16:44 node3 Keepalived\_vrrp\[31993\]: VRRP\_Instance(VI_1) Entering BACKUP STATE
Oct 30 00:16:44 node3 Keepalived\_vrrp\[31993\]: VRRP\_Instance(VI_1) removing protocol VIPs.
Oct 30 00:16:44 node3 Keepalived_healthcheckers\[31992\]: Netlink reflector reports IP 192.168.254.200 removed  #->该VIP已转移到slave.
.......
.......
\[root@master keepalived\]# ip a   #->VIP 已转移到slave
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid\_lft forever preferred\_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER\_UP> mtu 1500 qdisc pfifo\_fast state UP qlen 1000
link/ether 00:0c:29:5d:7d:94 brd ff:ff:ff:ff:ff:ff
inet 192.168.254.47/24 brd 192.168.254.255 scope global eth0
inet6 fe80::20c:29ff:fe5d:7d94/64 scope link
valid\_lft forever preferred\_lft forever
\[root@master keepalived\]#
slave:
\[root@slave keepalived\]# ip a
.......
.......
inet 192.168.254.200/32 scope global eth0   #->VIP 已转移过来
.......
.......
\[root@slave keepalived\]#

  **至此自写监测脚本，完成维护模式切换，已经完成；下面来解决最后一个问题：** **4.3.****如何在****Keepalived****主从切换时向管理员发送通知邮件？** (1)Keepalived通知脚本进阶示例：

下面的脚本可以接受选项，其中：
-s, --service SERVICE,...：指定服务脚本名称，当状态切换时可自动启动、重启或关闭此服务；
-a, --address VIP: 指定相关虚拟路由器的VIP地址；
-m, --mode {mm|mb}：指定虚拟路由的模型，mm表示主主，mb表示主备；它们表示相对于同一种服务而方，其VIP的工作类型；
-n, --notify {master|backup|fault}：指定通知的类型，即vrrp角色切换的目标角色；
-h, --help：获取脚本的使用帮助；

#!/bin/bash
\# Author: Tux
\# description: An example of notify script
\# Usage: notify.sh -m|--mode {mm|mb} -s|--service SERVICE1,... -a|--address VIP&nbsp; -n|--notify {master|backup|falut} -h|--help

&nbsp;

helpflag=0
serviceflag=0
modeflag=0
addressflag=0
notifyflag=0

contact='2399447849@qq.com'&nbsp;&nbsp; #->指定联系人;可以有多个，用”,”分隔开来


<br>
Usage() {
  echo "Usage: notify.sh \[-m|--mode {mm|mb}\] \[-s|--service SERVICE1,...\] <-a|--address VIP>  <-n|--notify {master|backup|falut}>" 
  echo "Usage: notify.sh -h|--help"
}
########################################################################################
ParseOptions() {
  local I=1;
  if \[ $# -gt 0 \]; then
    while \[ $I -le $# \]; do
      case $1 in
	  -s|--service)
		\[ $# -lt 2 \] && return 3
 	    serviceflag=1
 		services=(\`echo $2|awk -F"," '{for(i=1;i<=NF;i++) print $i}'\`)
		shift 2 ;;
	  -h|--help)
 		helpflag=1
		return 0
        shift
		;;
	  -a|--address)
		\[ $# -lt 2 \] && return 3
	    addressflag=1
		vip=$2
		shift 2
		;;
	  -m|--mode)
		\[ $# -lt 2 \] && return 3
		mode=$2
		shift 2
		;;
	  -n|--notify)
		\[ $# -lt 2 \] && return 3
		notifyflag=1
		notify=$2
		shift 2
		;;
	  *)
		echo "Wrong options..."
		Usage
		return 7
		;;
       esac
    done
    return 0
  fi
}

#workspace=$(dirname $0)

RestartService() {
  if \[ ${#@} -gt 0 \]; then
    for I in $@; do
      if \[ -x /etc/rc.d/init.d/$I \]; then
        /etc/rc.d/init.d/$I restart
      else
        echo "$I is not a valid service..."
      fi
    done
  fi
}

StopService() {
  if \[ ${#@} -gt 0 \]; then
    for I in $@; do
      if \[ -x /etc/rc.d/init.d/$I \]; then
        /etc/rc.d/init.d/$I stop
      else
        echo "$I is not a valid service..."
      fi
    done
  fi
}


Notify() {
    mailsubject="\`hostname\` to be $1: $vip floating"
    mailbody="\`date '+%F %H:%M:%S'\`, vrrp transition, \`hostname\` changed to be $1."
    echo $mailbody | mail -s "$mailsubject" $contact
}


\# Main Function
ParseOptions $@
\[ $? -ne 0 \] && Usage && exit 5

\[ $helpflag -eq 1 \] && Usage && exit 0

if \[ $addressflag -ne 1 -o $notifyflag -ne 1 \]; then
  Usage
  exit 2
fi

mode=${mode:-mb}

case $notify in
'master')
  if \[ $serviceflag -eq 1 \]; then
      RestartService ${services\[*\]}
  fi
  Notify master
  ;;
'backup')
  if \[ $serviceflag -eq 1 \]; then
    if \[ "$mode" == 'mb' \]; then
      StopService ${services\[*\]}
    else
      RestartService ${services\[*\]}
    fi
  fi
  Notify backup
  ;;
'fault')
  Notify fault
  ;;
*)
  Usage
  exit 4
  ;;
esac

(2) 在keepalived.conf配置文件中，其调用方法如下所示：

notify\_master "/etc/keepalived/notify.sh -n master -a VIP\_address"
notify\_backup "/etc/keepalived/notify.sh -n backup -a VIP\_address"
notify\_fault "/etc/keepalived/notify.sh -n fault -a VIP\_address"

(3)修改master/slave 的keepalived配置文件：、  

master:
\[root@master ~\]# vim /etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {
notification_email {
2399447849@qq.com
}

notification\_email\_from root
smtp_server 127.0.0.1
smtp\_connect\_timeout 30
router\_id LVS\_DEVEL
}

 

vrrp\_script chk\_schedown {
script "\[ -e /etc/keepalived/down \] && exit 1 || exit 0"
interval 1
weight -5
fall 2
rise 1
}


vrrp\_instance VI\_1 {
state MASTER
interface eth0
virtual\_router\_id 60
priority 101
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}

virtual_ipaddress {
192.168.254.200
}

track_script {
     chk_schedown
}

#->增加以下三行(注：在slave上面也一样添加这三行，此处略)
    notify_master "/etc/keepalived/notify.sh -n master -a 192.168.254.200"
    notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.254.200"
    notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.254.200"
}

 

virtual_server 192.168.254.200 80 {
delay_loop 6
lb_algo rr
lb_kind DR
nat_mask 255.255.255.0
#persistence_timeout 50
protocol TCP
 

real_server 192.168.254.45 80  {
weight 1
HTTP_GET {
    url {
      path /
      status_code 200
}

connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
}

}

real_server 192.168.254.46 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb\_get\_retry 3
delay\_before\_retry 3
}
}

sorry_server 127.0.0.1 80
}

  (4)添加脚本： 讲上述的脚本添加至master和slave的/etc/keepalived/目录下(注意权限)：

\[root@master ~\]# ll /etc/keepalived
总用量 8
-rw-r--r--. 1 root root 1748 10月 30 17:08 keepalived.conf
-rwxr-xr-x. 1 root root 2380 10月 30 00:57 notify.sh
\[root@master ~\]#
 
#->复制至slave
\[root@master ~\]# scp /etc/keepalived/notify.sh 192.168.254.48:/etc/keepalived/

(5)测试一下脚本可用性：

\[root@slave keepalived\]# ./notify.sh --help
Usage: notify.sh \[-m|--mode {mm|mb}\] \[-s|--service SERVICE1,...\] <-a|--address VIP>  <-n|--notify {master|backup|falut}>
Usage: notify.sh -h|--help
\[root@slave keepalived\]# ./notify.sh -m mb -a 2.2.2.2 -n master
\[root@slave keepalived\]#

查看邮件: [![12](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/12.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/12.png) 在模拟故障时重启一下keepalived,以免前面的实验造成影响。 注：现在已经可以成功收到邮件，通知脚本可用； (6)故障模拟： <1>先重启主备keepalived服务

master:
\[root@master ~\]# service keepalived restart
停止 keepalived：                                          \[确定\]
正在启动 keepalived：                                      \[确定\]
\[root@master ~\]#
slave:
\[root@slave ~\]# service keepalived restart
停止 keepalived：                                          \[确定\]
正在启动 keepalived：                                      \[确定\]
\[root@slave ~\]#

<2>正常情况下此时VIP在master上面

master:
\[root@master ~\]# ip a
.......
.......
inet 192.168.254.200/32 scope global eth0
.......
.......
\[root@master ~\]#

<3>在master的/etc/keepalived目录下 touch一个文件”down”

master:
\[root@master keepalived\]# touch down

<4>观察VIP 转移情况

slave:
\[root@slave ~\]# ip a
.......
.......
inet 192.168.254.200/32 scope global eth0
.......
.......
\[root@slave ~\]#

<5>结果查看—>邮件收取 [

13](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/13.png)](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/13.png) <6>Client访问测试 [![14](https://qcloud.coding.net/u/guomaoqiu/p/guomaoqiu/git/raw/master/uploads/2014/10/14.png)

从上可以看到，在keepalived主备切换时，不仅能够发送邮件，而且访问服务也没有问题；至此Lvs+Keepalived的基本应用实验演示完毕！

LVS+Keepalived

谢谢大爷~