RHEL 5.4下部署LVS(DR)+keepalived实现高性能高可用负载均衡

一、简介

LVS是Linux Virtual Server的简写，意即Linux虚拟服务器，是一个虚拟的服务器集群系统。本项目在1998年5月由章文嵩博士成立，是中国国内最早出现的自由软件项目之一。

目前有三种IP负载均衡技术（VS/NAT、VS/TUN和VS/DR）；十种调度算法（rr|wrr|lc|wlc|lblc|lblcr|dh|sh|sed|nq）。

Keepalived在这里主要用作RealServer的健康状态检查以及LoadBalance主机和BackUP主机之间failover的实现。

LVS+Keepalived能实现的功能：利用LVS控制器主备模式避免单点故障以及自动删除故障WEB服务器结点并当它恢复后再自动添加到群集中。

拓扑图：

二、系统环境

系统平台：RHEL 5.4

LVS版本：ipvsadm-1.24

keepalived版本：keepalived-1.1.15

三、安装

1、在两台LVS Server上安装lvs、keepalived软件（LVS_MASTER & LVS_BACKUP）
lvs install -------------

[root@LVS-MASTER ~]# wget http://www.linuxvirtualserver.org/software/kernel-2.6/ipvsadm-1.24.tar.gz

[root@LVS-MASTER ~]# ln -s /usr/src/kernels/2.6.18-194.el5-i686//usr/src/linux/

[root@LVS-MASTER ~]# tar zxvf ipvsadm-1.24.tar.gz

[root@LVS-MASTER ~]# cd ipvsadm-1.24

[root@LVS-MASTER ipvsadm-1.24]# make && make install

Keepalived install -------------

[root@LVS-MASTER ~]# wget http://www.keepalived.org/software/keepalived-1.1.15.tar.gz

[root@LVS-MASTER ~]# tar zxvf keepalived-1.1.15.tar.gz

[root@LVS-MASTER ~]# cd keepalived-1.1.15

[root@LVS-MASTER keepalived-1.1.15]# ./configure && make && make install

######### 将keepalived做成启动服务，方便管理##########

[root@LVS-MASTER ~]# cp /usr/local/etc/rc.d/init.d/keepalived /etc/init.d/

[root@LVS-MASTER ~]# cp /usr/local/etc/sysconfig/keepalived /etc/sysconfig/

[root@LVS-MASTER ~]# mkdir /etc/keepalived/

[root@LVS-MASTER ~]# cp /usr/local/etc/keepalived/keepalived.conf /etc/keepalived/

[root@LVS-MASTER ~]# cp /usr/local/sbin/keepalived /usr/sbin/

[root@LVS-MASTER ~]# service keepalived start | stop

2、keepalived的配置

###################  MASTER ###################

 ! Configuration File for keepalived

global_defs {
   notification_email {
     [email protected]               #设置报警邮件地址，可以设置多个，每行一个。
     [email protected]                 #需开启本机的sendmail服务
   }
   notification_email_from keepalived@localhost         #设置邮件的发送地址
   smtp_server 127.0.0.1                                #设置smtp server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 51      #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 100              #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
        10.0.0.148
    }
}

virtual_server 10.0.0.148 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
    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，
                                    #那么接下来的操作会被分发到另外的节点，但是如果用户一直在操作动态页面，则不受50秒的时间限制
    protocol TCP                    #指定转发协议类型，有TCP和UDP两种

    real_server 10.0.0.137 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 3                    #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
        TCP_CHECK {                 #realserver的状态检测设置部分，单位是秒
            connect_timeout 10      #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
            connect_port 80
        }
    }
    real_server 10.0.0.139 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 10
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}

#################### BACKUP ###################

! Configuration File for keepalived

global_defs {
   notification_email {
     [email protected]
     [email protected]
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}

vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 51
    priority 99
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.0.0.148
    }
}

virtual_server 10.0.0.148 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    persistence_timeout 50
    protocol TCP

    real_server 10.0.0.137 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 10
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 10.0.0.139 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 10
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}

3、在两台Web Server上执行realserver.sh脚本，为lo:0绑定VIP地址10.0.0.148、抑制ARP广播。

[root@web1 ~]# cat realserver.sh

#!/bin/bash
#description: Config realserver

VIP=10.0.0.148
 
/etc/rc.d/init.d/functions
 
case "$1" in
start)
       /sbin/ifconfig lo:0 $VIP netmask 255.255.255.255 broadcast $VIP
       /sbin/route add -host $VIP dev lo:0
       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
       sysctl -p >/dev/null 2>&1
       echo "RealServer Start OK"
       ;;
stop)
       /sbin/ifconfig lo:0 down
       /sbin/route del $VIP >/dev/null 2>&1
       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
       echo "RealServer Stoped"
       ;;
*)
       echo "Usage: $0 {start|stop}"
       exit 1
esac
 
exit 0

[root@web1 ~]# sh realserver.sh start

4、分别在LVS-MASTER、LVS-BACKUP上执行service keepalived start启动keepalived就可实现负载均衡及高可用集群了；

[root@LVS-MASTER keepalived]# service keepalived start

[root@LVS-MASTER keepalived]# ipvsadm -L
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.0.0.148:http rr persistent 50
  -> 10.0.0.139:http              Route   3      0          0         
  -> 10.0.0.137:http              Route   3      0          0         
[root@LVS-MASTER keepalived]#

查看日志/var/log/messages

[root@LVS-MASTER keepalived]# tail -n 30/var/log/messages
Apr 19 03:59:58 LVS-MASTER avahi-daemon[2845]: Withdrawing address record for10.0.0.148 on eth0.
Apr 19 04:00:19 LVS-MASTER Keepalived: Starting Keepalived v1.2.1 (04/17,2012)
Apr 19 04:00:19 LVS-MASTER Keepalived: Starting Healthcheck child process,pid=5960
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Netlink reflector reportsIP 10.0.0.136 added
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Netlink reflector reportsIP 10.0.0.159 added
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Registering Kernelnetlink reflector
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Registering Kernelnetlink command channel
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Opening file '/etc/keepalived/keepalived.conf'.
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Configuration is using :12257 Bytes
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Using LinkWatch kernelnetlink reflector...
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Activating healtcheckerfor service [10.0.0.137:80]
Apr 19 04:00:19 LVS-MASTER Keepalived_healthcheckers: Activating healtcheckerfor service [10.0.0.139:80]
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Netlink reflector reports IP10.0.0.136 added
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Netlink reflector reports IP10.0.0.159 added
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Registering Kernel netlinkreflector
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Registering Kernel netlink commandchannel
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Registering gratutious ARP sharedchannel
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Opening file'/etc/keepalived/keepalived.conf'.
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Configuration is using : 36698Bytes
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: Using LinkWatch kernel netlinkreflector...
Apr 19 04:00:19 LVS-MASTER Keepalived_vrrp: VRRP sockpool: [ifindex(2),proto(112), fd(11,12)]
Apr 19 04:00:19 LVS-MASTER Keepalived: Starting VRRP child process, pid=5961
Apr 19 04:00:20 LVS-MASTER Keepalived_vrrp: VRRP_Instance(VI_1) Transition toMASTER STATE
Apr 19 04:00:21 LVS-MASTER Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTERSTATE
Apr 19 04:00:21 LVS-MASTER Keepalived_vrrp: VRRP_Instance(VI_1) settingprotocol VIPs.
Apr 19 04:00:21 LVS-MASTER Keepalived_vrrp: VRRP_Instance(VI_1) Sendinggratuitous ARPs on eth0 for 10.0.0.148
Apr 19 04:00:21 LVS-MASTER Keepalived_healthcheckers: Netlink reflector reportsIP 10.0.0.136 added
Apr 19 04:00:21 LVS-MASTER avahi-daemon[2845]: Registering new address recordfor 10.0.0.148 on eth0.
Apr 19 04:00:21 LVS-MASTER Keepalived_vrrp: Netlink reflector reports IP10.0.0.136 added
Apr 19 04:00:26 LVS-MASTER Keepalived_vrrp: VRRP_Instance(VI_1) Sendinggratuitous ARPs on eth0 for 10.0.0.148
[root@LVS-MASTER keepalived]#

四、测试

接下来做高可用性测试&故障切换测试......

####高可用性测试####

模拟故障，将LVS-MASTER上的keepalived服务停掉，然后观察LVS-BACKUP上的日志，信息如下

[root@LVS-BACKUP keepalived]# tail -f /var/log/messages
Apr 19 03:33:04 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1)Transition to MASTER STATE
Apr 19 03:33:05 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTERSTATE
Apr 19 03:33:05 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) settingprotocol VIPs.
Apr 19 03:33:05 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) Sendinggratuitous ARPs on eth0 for 10.0.0.148
Apr 19 03:33:05 LVS-BACKUP Keepalived_vrrp: Netlink reflector reports IP10.0.0.157 added
Apr 19 03:33:05 LVS-BACKUP Keepalived_healthcheckers: Netlink reflector reportsIP 10.0.0.157 added
Apr 19 03:33:05 LVS-BACKUP avahi-daemon[2825]: Registering new address recordfor 10.0.0.148 on eth0.
Apr 19 03:33:10 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) Sendinggratuitous ARPs on eth0 for 10.0.0.148

从日志中可知，主机出现故障后，备机立刻检测到，此时备机变为MASTER角色，并且接管了主机的虚拟IP资源，最后将虚拟IP绑定在etho设备上。

将LVS-MASTER 上的keepalived服务开启后，LVS-BACKUP的日志状态。

Apr 19 03:34:02 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) Received higher prio advert
Apr 19 03:34:02 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1)Entering BACKUP STATE
Apr 19 03:34:02 LVS-BACKUP Keepalived_vrrp: VRRP_Instance(VI_1) removing protocol VIPs.
Apr 19 03:34:02 LVS-BACKUP Keepalived_vrrp: Netlink reflector reports IP 10.0.0.157 removed
Apr 19 03:34:02 LVS-BACKUP Keepalived_healthcheckers: Netlink reflector reports IP 10.0.0.157 removed
Apr 19 03:34:02 LVS-BACKUP avahi-daemon[2825]: Withdrawing address record for 10.0.0.148 on eth0.

从日志可知，备机在检测到主机重新恢复正常后，释放了虚拟IP资源重新成为BACKUP角色

####故障切换测试####

故障切换是测试当某个节点出现故障后，Keepalived监制模块是否能及时发现然后屏蔽故障节点，同时将服务器转移到正常节点来执行。

将web2节点服务停掉，假设这个节点出现故障，然后主、备机日志信息如下

Apr 19 03:35:04 LVS-MASTER Keepalived_healthcheckers:TCP connection to [10.0.0.139:80] failed!!!
Apr 19 03:35:04 LVS-MASTER Keepalived_healthcheckers:Removing service [10.0.0.139:80] from VS [10.0.0.148:80]
Apr 19 03:35:04 LVS-MASTER Keepalived_healthcheckers: Remote SMTP server [127.0.0.1:25] connected.

[root@LVS-MASTER keepalived]# ipvsadm -L
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.0.0.148:http rr persistent 50
  -> 10.0.0.137:http              Route   3      0          0         
[root@LVS-MASTER keepalived]#

从以上可以看出，Keepalived监控模块检测到10.0.0.139这台主机出现故障后，将些web2从集群系统中剔除掉了。 此时访问http://10.0.0.148只能看到web1了)

重新启动web2节点的服务，日志信息如下：

Apr 19 03:38:22 LVS-MASTER Keepalived_healthcheckers:TCP connection to [10.0.0.139:80] success.
Apr 19 03:38:22 LVS-MASTER Keepalived_healthcheckers:Adding service [10.0.0.139:80] to VS [10.0.0.148:80]
Apr 19 03:38:22 LVS-MASTER Keepalived_healthcheckers: Remote SMTP server [127.0.0.1:25] connected.

[root@LVS-MASTERkeepalived]# ipvsadm -L
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.0.0.148:http rr persistent 50
  -> 10.0.0.139:http              Route   3      0          0         
  -> 10.0.0.137:http              Route   3      0          0         
[root@LVS-MASTER keepalived]#

Keepalived监控模块检测到10.0.0.139这台主机恢复正常后，又将此节点加入集群系统中，再次访问就可以访问到web2页面了）