LVS+Keepalived 高可用环境部署记录(主主和主从模式)

之前的文章介绍了LVS负载均衡-基础知识梳理, 下面记录下LVS+Keepalived高可用环境部署梳理(主主和主从模式)的操作流程:

一、LVS+Keepalived主从热备的高可用环境部署

1)环境准备

LVS_Keepalived_Master      182.148.15.237
LVS_Keepalived_Backup      182.148.15.236
Real_Server1               182.148.15.233
Real_Server2               182.148.15.238
VIP                        182.148.15.239
 
系统版本都是centos6.8

特别注意:
Director Server与Real Server必须有一块网卡连在同一物理网段上!否则lvs会转发失败!
在远程telnet vip port会报错:
"telnet: connect to address *.*.*.*: No route to host"

基本的网络拓扑图如下:

LVS+Keepalived 高可用环境部署记录(主主和主从模式)

2)LVS_keepalived_Master和LVS_keepalived_Backup两台服务器上安装配置LVS和keepalived的操作记录:

1)关闭 SElinux、配置防火墙(在LVS_Keepalived_Master 和 LVS_Keepalived_Backup两台机器上都要操作)
[root@LVS_Keepalived_Master ~]# vim /etc/sysconfig/selinux
#SELINUX=enforcing                #注释掉
#SELINUXTYPE=targeted             #注释掉
SELINUX=disabled                  #增加
   
[root@LVS_Keepalived_Master ~]# setenforce 0      #临时关闭selinux。上面文件配置后,重启机器后就永久生效。
   
注意下面182.148.15.0/24是服务器的公网网段,192.168.1.0/24是服务器的私网网段
一定要注意:加上这个组播规则后,MASTER和BACKUP故障时,才能实现VIP资源的正常转移。其故障恢复后,VIP也还会正常转移回来。
[root@LVS_Keepalived_Master ~]# vim /etc/sysconfig/iptables 
.......
-A INPUT -s 182.148.15.0/24 -d 224.0.0.18 -j ACCEPT      #允许组播地址通信。
-A INPUT -s 192.168.1.0/24 -d 224.0.0.18 -j ACCEPT
-A INPUT -s 182.148.15.0/24 -p vrrp -j ACCEPT            #允许 VRRP(虚拟路由器冗余协)通信
-A INPUT -s 192.168.1.0/24 -p vrrp -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 80 -j ACCEPT
   
[root@LVS_Keepalived_Master ~]# /etc/init.d/iptables restart
  
----------------------------------------------------------------------------------------------------------------------
2)LVS安装(在LVS_Keepalived_Master 和 LVS_Keepalived_Backup两台机器上都要操作)
需要安装以下软件包
[root@LVS_Keepalived_Master ~]# yum install -y libnl* popt*
  
查看是否加载lvs模块
[root@LVS_Keepalived_Master src]# modprobe -l |grep ipvs
  
下载并安装LVS
[root@LVS_Keepalived_Master ~]# cd /usr/local/src/
[root@LVS_Keepalived_Master src]# wget http://www.linuxvirtualserver.org/software/kernel-2.6/ipvsadm-1.26.tar.gz
解压安装
[root@LVS_Keepalived_Master src]# ln -s /usr/src/kernels/2.6.32-431.5.1.el6.x86_64/ /usr/src/linux
[root@LVS_Keepalived_Master src]# tar -zxvf ipvsadm-1.26.tar.gz
[root@LVS_Keepalived_Master src]# cd ipvsadm-1.26
[root@LVS_Keepalived_Master ipvsadm-1.26]# make && make install
   
LVS安装完成,查看当前LVS集群
[root@LVS_Keepalived_Master ipvsadm-1.26]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
  
----------------------------------------------------------------------------------------------------------------------
3)编写LVS启动脚本/etc/init.d/realserver(在Real_Server1 和Real_Server2上都要操作,realserver脚本内容是一样的)
[root@Real_Server1 ~]# vim /etc/init.d/realserver
#!/bin/sh
VIP=182.148.15.239
. /etc/rc.d/init.d/functions
   
case "$1" in
# 禁用本地的ARP请求、绑定本地回环地址
start)
    /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/sysctl -p >/dev/null 2>&1
    /sbin/ifconfig lo:0 $VIP netmask 255.255.255.255 up     #在回环地址上绑定VIP,设定掩码,与Direct Server(自身)上的IP保持通信
    /sbin/route add -host $VIP dev lo:0
    echo "LVS-DR real server starts successfully.\n"
    ;;
stop)
    /sbin/ifconfig lo:0 down
    /sbin/route del $VIP >/dev/null 2>&1
    echo "1" >/proc/sys/net/ipv4/conf/lo/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/lo/arp_announce
    echo "1" >/proc/sys/net/ipv4/conf/all/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/all/arp_announce
echo "LVS-DR real server stopped.\n"
    ;;
status)
    isLoOn=`/sbin/ifconfig lo:0 | grep "$VIP"`
    isRoOn=`/bin/netstat -rn | grep "$VIP"`
    if [ "$isLoON" == "" -a "$isRoOn" == "" ]; then
        echo "LVS-DR real server has run yet."
    else
        echo "LVS-DR real server is running."
    fi
    exit 3
    ;;
*)
    echo "Usage: $0 {start|stop|status}"
    exit 1
esac
exit 0
  
将lvs脚本加入开机自启动
[root@Real_Server1 ~]# chmod +x /etc/init.d/realserver
[root@Real_Server1 ~]# echo "/etc/init.d/realserver start" >> /etc/rc.d/rc.local
  
启动LVS脚本(注意:如果这两台realserver机器重启了,一定要确保service realserver start  启动了,即lo:0本地回环上绑定了vip地址,否则lvs转发失败!)
[root@Real_Server1 ~]# service realserver start
LVS-DR real server starts successfully.\n
  
查看Real_Server1服务器,发现VIP已经成功绑定到本地回环口lo上了
[root@Real_Server1 ~]# ifconfig
eth0      Link encap:Ethernet  HWaddr 52:54:00:D1:27:75
          inet addr:182.148.15.233  Bcast:182.148.15.255  Mask:255.255.255.224
          inet6 addr: fe80::5054:ff:fed1:2775/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:309741 errors:0 dropped:0 overruns:0 frame:0
          TX packets:27993954 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:37897512 (36.1 MiB)  TX bytes:23438654329 (21.8 GiB)
  
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
  
lo:0      Link encap:Local Loopback
          inet addr:182.148.15.239  Mask:255.255.255.255
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
  
----------------------------------------------------------------------------------------------------------------------
4)安装Keepalived(LVS_Keepalived_Master 和 LVS_Keepalived_Backup两台机器都要操作)
[root@LVS_Keepalived_Master ~]# yum install -y openssl-devel
[root@LVS_Keepalived_Master ~]# cd /usr/local/src/
[root@LVS_Keepalived_Master src]# wget http://www.keepalived.org/software/keepalived-1.3.5.tar.gz
[root@LVS_Keepalived_Master src]# tar -zvxf keepalived-1.3.5.tar.gz
[root@LVS_Keepalived_Master src]# cd keepalived-1.3.5
[root@LVS_Keepalived_Master keepalived-1.3.5]# ./configure --prefix=/usr/local/keepalived
[root@LVS_Keepalived_Master keepalived-1.3.5]# make && make install
         
[root@LVS_Keepalived_Master keepalived-1.3.5]# cp /usr/local/src/keepalived-1.3.5/keepalived/etc/init.d/keepalived /etc/rc.d/init.d/
[root@LVS_Keepalived_Master keepalived-1.3.5]# cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
[root@LVS_Keepalived_Master keepalived-1.3.5]# mkdir /etc/keepalived/
[root@LVS_Keepalived_Master keepalived-1.3.5]# cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/
[root@LVS_Keepalived_Master keepalived-1.3.5]# cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
[root@LVS_Keepalived_Master keepalived-1.3.5]# echo "/etc/init.d/keepalived start" >> /etc/rc.local
   
[root@LVS_Keepalived_Master keepalived-1.3.5]# chmod +x /etc/rc.d/init.d/keepalived      #添加执行权限
[root@LVS_Keepalived_Master keepalived-1.3.5]# chkconfig keepalived on                   #设置开机启动
[root@LVS_Keepalived_Master keepalived-1.3.5]# service keepalived start                   #启动
[root@LVS_Keepalived_Master keepalived-1.3.5]# service keepalived stop                    #关闭
[root@LVS_Keepalived_Master keepalived-1.3.5]# service keepalived restart                 #重启
  
----------------------------------------------------------------------------------------------------------------------
5)接着配置LVS+Keepalived配置

现在LVS_Keepalived_Master和LVS_Keepalived_Backup两台机器上打开ip_forward转发功能
[root@LVS_Keepalived_Master ~]# echo "1" > /proc/sys/net/ipv4/ip_forward
  
LVS_Keepalived_Master机器上的keepalived.conf配置:
[root@LVS_Keepalived_Master ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
   
global_defs {
   router_id LVS_Master
}
   
vrrp_instance VI_1 {
    state MASTER               #指定instance初始状态,实际根据优先级决定.backup节点不一样
    interface eth0             #虚拟IP所在网
    virtual_router_id 51       #VRID,相同VRID为一个组,决定多播MAC地址
    priority 100               #优先级,另一台改为90.backup节点不一样
    advert_int 1               #检查间隔
    authentication {
        auth_type PASS         #认证方式,可以是pass或ha
        auth_pass 1111         #认证密码
    }
    virtual_ipaddress {
        182.148.15.239         #VIP
    }
}
   
virtual_server 182.148.15.239 80 {
    delay_loop 6               #服务轮询的时间间隔
    lb_algo wrr                #加权轮询调度,LVS调度算法 rr|wrr|lc|wlc|lblc|sh|sh
    lb_kind DR                 #LVS集群模式 NAT|DR|TUN,其中DR模式要求负载均衡器网卡必须有一块与物理网卡在同一个网段
    #nat_mask 255.255.255.0
    persistence_timeout 50     #会话保持时间
    protocol TCP              #健康检查协议
   
    ## Real Server设置,80就是连接端口
    real_server 182.148.15.233 80 {
        weight 3  ##权重
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
  
启动keepalived
[root@LVS_Keepalived_Master ~]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
  
[root@LVS_Keepalived_Master ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:68:dc:b6 brd ff:ff:ff:ff:ff:ff
    inet 182.48.115.237/27 brd 182.48.115.255 scope global eth0
    inet 182.48.115.239/32 scope global eth0
    inet6 fe80::5054:ff:fe68:dcb6/64 scope link
       valid_lft forever preferred_lft forever
  
注意此时网卡的变化,可以看到虚拟网卡已经分配到了realserver上。
此时查看LVS集群状态,可以看到集群下有两个Real Server,调度算法,权重等信息。ActiveConn代表当前Real Server的活跃连接数。
[root@LVS_Keepalived_Master ~]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  182.48.115.239:80 wrr persistent 50
  -> 182.48.115.233:80            Route   3      0          0       
  -> 182.48.115.238:80            Route   3      0          0 
  
-------------------------------------------------------------------------
LVS_Keepalived_Backup机器上的keepalived.conf配置:
[root@LVS_Keepalived_Backup ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
   
global_defs {
   router_id LVS_Backup
}
   
vrrp_instance VI_1 {
    state BACKUP           
    interface eth0          
    virtual_router_id 51    
    priority 90            
    advert_int 1           
    authentication {
        auth_type PASS      
        auth_pass 1111      
    }
    virtual_ipaddress {
        182.148.15.239      
    }
}
   
virtual_server 182.148.15.239 80 {
    delay_loop 6           
    lb_algo wrr            
    lb_kind DR              
   
    persistence_timeout 50  
    protocol TCP          
   
    real_server 182.148.15.233 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
  
[root@LVS_Keepalived_Backup ~]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
  
查看LVS_Keepalived_Backup机器上,发现VIP默认在LVS_Keepalived_Master机器上,只要当LVS_Keepalived_Backup发生故障时,VIP资源才会飘到自己这边来。
[root@LVS_Keepalived_Backup ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:7c:b8:f0 brd ff:ff:ff:ff:ff:ff
    inet 182.48.115.236/27 brd 182.48.115.255 scope global eth0
    inet 182.48.115.239/27 brd 182.48.115.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe7c:b8f0/64 scope link
       valid_lft forever preferred_lft forever
[root@LVS_Keepalived_Backup ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  182.48.115.239:80 wrr persistent 50
  -> 182.48.115.233:80            Route   3      0          0       
  -> 182.48.115.238:80            Route   3      0          0

3)后端两台Real Server上的操作

在两台Real Server上配置好nginx,nginx安装配置过程省略。
分别在两台Real Server上配置两个域名www.wangshibo.com和www.guohuihui.com。

在LVS_Keepalived_Master 和 LVS_Keepalived_Backup两台机器上要能正常访问这两个域名
[root@LVS_Keepalived_Master ~]# curl http://www.wangshibo.com
this is page of Real_Server1:182.148.15.238 www.wangshibo.com
[root@LVS_Keepalived_Master ~]# curl http://www.guohuihui.com
this is page of Real_Server2:182.148.15.238 www.guohuihui.com

[root@LVS_Keepalived_Backup ~]# curl http://www.wangshibo.com
this is page of Real_Server1:182.148.15.238 www.wangshibo.com
[root@LVS_Keepalived_Backup ~]# curl http://www.guohuihui.com
this is page of Real_Server2:182.148.15.238 www.guohuihui.com


关闭182.148.15.238这台机器(即Real_Server2)的nginx,发现对应域名的请求就会到Real_Server1上
[root@Real_Server2 ~]# /usr/local/nginx/sbin/nginx -s stop
[root@Real_Server2 ~]# lsof -i:80
[root@Real_Server2 ~]# 

再次在LVS_Keepalived_Master 和 LVS_Keepalived_Backup两台机器*问这两个域名,就会发现已经负载到Real_Server1上了
[root@LVS_Keepalived_Master ~]# curl http://www.wangshibo.com
this is page of Real_Server1:182.148.15.233 www.wangshibo.com
[root@LVS_Keepalived_Master ~]# curl http://www.guohuihui.com
this is page of Real_Server1:182.148.15.233 www.guohuihui.com

[root@LVS_Keepalived_Backup ~]# curl http://www.wangshibo.com
this is page of Real_Server1:182.148.15.233 www.wangshibo.com
[root@LVS_Keepalived_Backup ~]# curl http://www.guohuihui.com
this is page of Real_Server1:182.148.15.233 www.guohuihui.com

另外,设置这两台Real Server的iptables,让其80端口只对前面的两个vip资源开放
[root@Real_Server1 ~]# vim /etc/sysconfig/iptables
......
-A INPUT -s 182.148.15.239 -m state --state NEW -m tcp -p tcp --dport 80 -j ACCEPT
[root@Real_Server1 ~]# /etc/init.d/iptables restart

4)测试
将www.wangshibo.com和www.guohuihui.com测试域名解析到VIP:182.148.15.239,然后在浏览器里是可以正常访问的。

LVS+Keepalived 高可用环境部署记录(主主和主从模式)

LVS+Keepalived 高可用环境部署记录(主主和主从模式)

1)测试LVS功能(上面Keepalived的lvs配置中,自带了健康检查,当后端服务器的故障出现故障后会自动从lvs集群中踢出,当故障恢复后,再自动加入到集群中)
先查看当前LVS集群,如下:发现后端两台Real Server的80端口都运行正常
[root@LVS_Keepalived_Master ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  182.148.15.239:80 wrr persistent 50
  -> 182.148.15.233:80            Route   3      0          0        
  -> 182.148.15.238:80            Route   3      0          0
 
现在测试关闭一台Real Server,比如Real_Server2
[root@Real_Server2 ~]# /usr/local/nginx/sbin/nginx -s stop
 
过一会儿再次查看当前LVS集群,如下:发现Real_Server2已经被踢出当前LVS集群了
[root@LVS_Keepalived_Master ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  182.148.15.239:80 wrr persistent 50
  -> 182.148.15.233:80            Route   3      0          0
 
最后重启Real_Server2的80端口,发现LVS集群里又再次将其添加进来了
[root@Real_Server2 ~]# /usr/local/nginx/sbin/nginx
 
[root@LVS_Keepalived_Master ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  182.148.15.239:80 wrr persistent 50
  -> 182.148.15.233:80            Route   3      0          0        
  -> 182.148.15.238:80            Route   3      0          0  
 
以上测试中,http://www.wangshibo.com和http://www.guohuihui.com域名访问都不受影响。
 
 
2)测试Keepalived心跳测试的高可用
默认情况下,VIP资源是在LVS_Keepalived_Master上
[root@LVS_Keepalived_Master ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:68:dc:b6 brd ff:ff:ff:ff:ff:ff
    inet 182.148.15.237/27 brd 182.148.15.255 scope global eth0
    inet 182.148.15.239/32 scope global eth0
    inet 182.148.15.239/27 brd 182.148.15.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe68:dcb6/64 scope link
       valid_lft forever preferred_lft forever
 
然后关闭LVS_Keepalived_Master的keepalived,发现VIP就会转移到LVS_Keepalived_Backup上。
[root@LVS_Keepalived_Master ~]# /etc/init.d/keepalived stop
Stopping keepalived:                                       [  OK  ]
[root@LVS_Keepalived_Master ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:68:dc:b6 brd ff:ff:ff:ff:ff:ff
    inet 182.148.15.237/27 brd 182.148.15.255 scope global eth0
    inet 182.148.15.239/27 brd 182.148.15.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe68:dcb6/64 scope link
       valid_lft forever preferred_lft forever
 
查看系统日志,能查看到LVS_Keepalived_Master的VIP的移动信息
[root@LVS_Keepalived_Master ~]# tail -f /var/log/messages
.............
May  8 10:19:36 Haproxy_Keepalived_Master Keepalived_healthcheckers[20875]: TCP connection to [182.148.15.233]:80 failed.
May  8 10:19:39 Haproxy_Keepalived_Master Keepalived_healthcheckers[20875]: TCP connection to [182.148.15.233]:80 failed.
May  8 10:19:39 Haproxy_Keepalived_Master Keepalived_healthcheckers[20875]: Check on service [182.148.15.233]:80 failed after 1 retry.
May  8 10:19:39 Haproxy_Keepalived_Master Keepalived_healthcheckers[20875]: Removing service [182.148.15.233]:80 from VS [182.148.15.239]:80
 
[root@LVS_Keepalived_Backup ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:7c:b8:f0 brd ff:ff:ff:ff:ff:ff
    inet 182.148.15.236/27 brd 182.148.15.255 scope global eth0
    inet 182.148.15.239/32 scope global eth0
    inet 182.148.15.239/27 brd 182.148.15.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe7c:b8f0/64 scope link
       valid_lft forever preferred_lft forever
 
 
接着再重新启动LVS_Keepalived_Master的keepalived,发现VIP又转移回来了
[root@LVS_Keepalived_Master ~]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
[root@LVS_Keepalived_Master ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 52:54:00:68:dc:b6 brd ff:ff:ff:ff:ff:ff
    inet 182.148.15.237/27 brd 182.148.15.255 scope global eth0
    inet 182.148.15.239/32 scope global eth0
    inet 182.148.15.239/27 brd 182.148.15.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe68:dcb6/64 scope link
       valid_lft forever preferred_lft forever
 
 
查看系统日志,能查看到LVS_Keepalived_Master的VIP转移回来的信息
[root@LVS_Keepalived_Master ~]# tail -f /var/log/messages
.............
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: Sending gratuitous ARP on eth0 for 182.148.15.239
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: VRRP_Instance(VI_1) Sending/queueing gratuitous ARPs on eth0 for 182.148.15.239
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: Sending gratuitous ARP on eth0 for 182.148.15.239
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: Sending gratuitous ARP on eth0 for 182.148.15.239
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: Sending gratuitous ARP on eth0 for 182.148.15.239
May  8 10:23:12 Haproxy_Keepalived_Master Keepalived_vrrp[5863]: Sending gratuitous ARP on eth0 for 182.148.15.239

二、LVS+Keepalived主主热备的高可用环境部署

主主环境相比于主从环境,区别只在于:
1)LVS负载均衡层需要两个VIP。比如182.148.15.239和182.148.15.235 
2)后端的realserver上要绑定这两个VIP到lo本地回环口上
3)Keepalived.conf的配置相比于上面的主从模式也有所不同
 
主主架构的具体配置如下:
1)编写LVS启动脚本(在Real_Server1 和Real_Server2上都要操作,realserver脚本内容是一样的)
 
由于后端realserver机器要绑定两个VIP到本地回环口lo上(分别绑定到lo:0和lo:1),所以需要编写两个启动脚本

[root@Real_Server1 ~]# vim /etc/init.d/realserver1
#!/bin/sh
VIP=182.148.15.239
. /etc/rc.d/init.d/functions
   
case "$1" in
 
start)
    /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/sysctl -p >/dev/null 2>&1
    /sbin/ifconfig lo:0 $VIP netmask 255.255.255.255 up   
    /sbin/route add -host $VIP dev lo:0
    echo "LVS-DR real server starts successfully.\n"
    ;;
stop)
    /sbin/ifconfig lo:0 down
    /sbin/route del $VIP >/dev/null 2>&1
    echo "1" >/proc/sys/net/ipv4/conf/lo/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/lo/arp_announce
    echo "1" >/proc/sys/net/ipv4/conf/all/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/all/arp_announce
echo "LVS-DR real server stopped.\n"
    ;;
status)
    isLoOn=`/sbin/ifconfig lo:0 | grep "$VIP"`
    isRoOn=`/bin/netstat -rn | grep "$VIP"`
    if [ "$isLoON" == "" -a "$isRoOn" == "" ]; then
        echo "LVS-DR real server has run yet."
    else
        echo "LVS-DR real server is running."
    fi
    exit 3
    ;;
*)
    echo "Usage: $0 {start|stop|status}"
    exit 1
esac
exit 0
 
 
[root@Real_Server1 ~]# vim /etc/init.d/realserver2
#!/bin/sh
VIP=182.148.15.235
. /etc/rc.d/init.d/functions
   
case "$1" in
 
start)
    /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/sysctl -p >/dev/null 2>&1
    /sbin/ifconfig lo:1 $VIP netmask 255.255.255.255 up    
    /sbin/route add -host $VIP dev lo:1
    echo "LVS-DR real server starts successfully.\n"
    ;;
stop)
    /sbin/ifconfig lo:1 down
    /sbin/route del $VIP >/dev/null 2>&1
    echo "1" >/proc/sys/net/ipv4/conf/lo/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/lo/arp_announce
    echo "1" >/proc/sys/net/ipv4/conf/all/arp_ignore
    echo "2" >/proc/sys/net/ipv4/conf/all/arp_announce
echo "LVS-DR real server stopped.\n"
    ;;
status)
    isLoOn=`/sbin/ifconfig lo:1 | grep "$VIP"`
    isRoOn=`/bin/netstat -rn | grep "$VIP"`
    if [ "$isLoON" == "" -a "$isRoOn" == "" ]; then
        echo "LVS-DR real server has run yet."
    else
        echo "LVS-DR real server is running."
    fi
    exit 3
    ;;
*)
    echo "Usage: $0 {start|stop|status}"
    exit 1
esac
exit 0
 
将lvs脚本加入开机自启动
[root@Real_Server1 ~]# chmod +x /etc/init.d/realserver1
[root@Real_Server1 ~]# chmod +x /etc/init.d/realserver2
[root@Real_Server1 ~]# echo "/etc/init.d/realserver1" >> /etc/rc.d/rc.local
[root@Real_Server1 ~]# echo "/etc/init.d/realserver2" >> /etc/rc.d/rc.local
  
启动LVS脚本
[root@Real_Server1 ~]# service realserver1 start
LVS-DR real server starts successfully.\n
 
[root@Real_Server1 ~]# service realserver2 start
LVS-DR real server starts successfully.\n
 
查看Real_Server1服务器,发现VIP已经成功绑定到本地回环口lo上了
[root@Real_Server1 ~]# ifconfig
eth0      Link encap:Ethernet  HWaddr 52:54:00:D1:27:75
          inet addr:182.148.15.233  Bcast:182.148.15.255  Mask:255.255.255.224
          inet6 addr: fe80::5054:ff:fed1:2775/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:309741 errors:0 dropped:0 overruns:0 frame:0
          TX packets:27993954 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:37897512 (36.1 MiB)  TX bytes:23438654329 (21.8 GiB)
  
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
  
lo:0      Link encap:Local Loopback
          inet addr:182.148.15.239  Mask:255.255.255.255
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
 
lo:1      Link encap:Local Loopback
          inet addr:182.148.15.235  Mask:255.255.255.255
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
 
 
2)Keepalived.conf的配置 
LVS_Keepalived_Master机器上的Keepalived.conf配置
先打开ip_forward路由转发功能
[root@LVS_Keepalived_Master ~]# echo "1" > /proc/sys/net/ipv4/ip_forward

[root@LVS_Keepalived_Master ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
   
global_defs {
   router_id LVS_Master
}
   
vrrp_instance VI_1 {
    state MASTER              
    interface eth0            
    virtual_router_id 51      
    priority 100              
    advert_int 1              
    authentication {
        auth_type PASS        
        auth_pass 1111        
    }
    virtual_ipaddress {
        182.148.15.239        
    }
}
   
vrrp_instance VI_2 {
    state BACKUP           
    interface eth0          
    virtual_router_id 52   
    priority 90            
    advert_int 1           
    authentication {
        auth_type PASS      
        auth_pass 1111      
    }
    virtual_ipaddress {
        182.148.15.235    
    }
}
 
virtual_server 182.148.15.239 80 {
    delay_loop 6              
    lb_algo wrr               
    lb_kind DR                
    #nat_mask 255.255.255.0
    persistence_timeout 50    
    protocol TCP             
   
 
    real_server 182.148.15.233 80 {
        weight 3 
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
 
 
virtual_server 182.148.15.235 80 {
    delay_loop 6              
    lb_algo wrr               
    lb_kind DR                
    #nat_mask 255.255.255.0
    persistence_timeout 50    
    protocol TCP             
   
 
    real_server 182.148.15.233 80 {
        weight 3 
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
 
 
 
LVS_Keepalived_Backup机器上的Keepalived.conf配置
[root@LVS_Keepalived_Backup ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
   
global_defs {
   router_id LVS_Backup
}
   
vrrp_instance VI_1 {
    state Backup             
    interface eth0            
    virtual_router_id 51      
    priority 90              
    advert_int 1              
    authentication {
        auth_type PASS        
        auth_pass 1111        
    }
    virtual_ipaddress {
        182.148.15.239        
    }
}
   
vrrp_instance VI_2 {
    state Master          
    interface eth0          
    virtual_router_id 52  
    priority 100           
    advert_int 1           
    authentication {
        auth_type PASS      
        auth_pass 1111      
    }
    virtual_ipaddress {
        182.148.15.235    
    }
}
 
virtual_server 182.148.15.239 80 {
    delay_loop 6              
    lb_algo wrr               
    lb_kind DR                
    #nat_mask 255.255.255.0
    persistence_timeout 50    
    protocol TCP             
   
 
    real_server 182.148.15.233 80 {
        weight 3 
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
 
 
virtual_server 182.148.15.235 80 {
    delay_loop 6              
    lb_algo wrr               
    lb_kind DR                
    #nat_mask 255.255.255.0
    persistence_timeout 50    
    protocol TCP             
   
 
    real_server 182.148.15.233 80 {
        weight 3 
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
    real_server 182.148.15.238 80 {
        weight 3
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}
 
 
 
其他验证操作和上面主从模式一样~~~
上一篇:LVS负载均衡下session共享的实现方式-持久化连接


下一篇:LVS负载均衡集群架构