LVS+Keepalived高可用负载均衡集群架构实验

162次阅读

共计 29819 个字符，预计需要花费 75 分钟才能阅读完成。

实验环境：
LVS 和 keepalived 介绍：
1.keepalived 的设计目标是构建高可用的 LVS 负载均衡群集，可以调用 ipvsadm 工具来创建虚拟服务器，管理服务器池，而不仅仅用来做双机热备。
使用 keepalived 构建 LVS 群集更加简便易用，主要优势体现在：
对 LVS 负载调度器实现热备切换，提高可用性；
对服务器池中的节点进行健康检查，自动移除失效节点，恢复后再重新加入。

2. 在基于 LVS+Keepalived 实现的 LVS 群集结构中，至少包括两台热备的负载调度器，两台以上的节点服务器，本例将以 DR 模式的 LVS 群集为基础，增加一台从负载调度器，使用 Keepalived 来实现主、从调度器的热备，从而构建建有负载均衡、高可用两种能力的 LVS 网站群集平台。

3. 通过 LVS+Keepalived 构建的 LVS 集群，LVS 负载均衡用户请求到后端 tomcat 服务器，Keepalived 的作用是检测 web 服务器的状态，如果有一台 web 服务器死机，或工作出现故障，Keepalived 将检测到，并将有故障的 web 服务器从系统中剔除，当 web 服务器工作正常后 Keepalived 自动将 web 服务器加入到服务器群中，这些工作全部自动完成，不需要人工干涉，需要人工做的只是修复故障的 web 服务器。

Keepalived 简介
什么是 Keepalived：keepalived 可以实现服务的高可用或热备，用来防止单点故障的问题；而 Keepalived 的核心 VRRP 协议，VRRP 协议主要实现了在路由器或三层交换机处的冗余；Keepalived 就是使用 VRRP 协议来实现高可用的；下面一起来看一下 Keepalived 的原理图：
LVS+Keepalived 高可用负载均衡集群架构实验

Keepalived 启动后会有三个进程：
父进程：内存管理，子进程管理
子进程：VRRP 子进程
子进程：healthchecker 子进程
由上图可知：
两个子进程都被系统 WatchDog 看管，两个子进程各自复杂自己的事，healthchecker 子进程复杂检查各自服务器的健康程度，例如 HTTP，LVS 等，如果 healthchecker 子进程检查到 MASTER 上服务不可用了，就会通知本机上的 VRRP 子进程，让它删除通告，并且去掉虚拟 IP，转换为 BACKUP 状态；

实验架构图：
LVS+Keepalived 高可用负载均衡集群架构实验
1 <BR>

实验过程：

HA 高可用集群构建前提：
1.proxy 和 proxy2 节点时间必须同步；
建议使用 ntp 协议进行；
参考博客：http://sohudrgon.blog.51cto.com/3088108/1598314

2、节点之间必须要通过主机名互相通信；
建议使用 hosts 文件；

通信中使用的名字必须与其节点为上“uname -n”命令展示出的名字保持一致；
[root@proxy ~]# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
172.16.0.1 server.magelinux.com server
172.16.31.52 proxy.stu31.com proxy
172.16.31.53 proxy2.stu31.com proxy2
172.16.31.50 tom1.stu31.com tom1
172.16.31.51 tom2.stu31.com tom2

3、节点之间彼此 root 用户能基于 ssh 密钥方式进行通信；
节点 proxy：
# ssh-keygen -t rsa -P “”
# ssh-copy-id -i .ssh/id_rsa.pub proxy2
节点 proxy2：
# ssh-keygen -t rsa -P “”
# ssh-copy-id -i .ssh/id_rsa.pub proxy

测试 ssh 无密钥通信：
123 [root@proxy ~]# date ; ssh proxy2 date
Sat Jan 17 15:29:26 CST 2015
Sat Jan 17 15:29:26 CST 2015

一. 配置安装 LVS+keepalived
1. 在 Master 与 Backup 服务器上分别安装 Ipvsadm、Keepalived 软件包、这里使用的是 yum 安装方式
# yum install -y keepalived ipvsadm

注释：这里安装 Ipvsadm 工具主要是为了查看 lvs 规则使用，不安装 ipvsadm 工具也能成功配置规则，但不方便；

2. 修改 keepalived 配置文件
keepalived 的文件路径 /etc/keepalived/keepalived.conf

修改 Master 节点 proxy 的主配置文件
[root@proxy ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {#设置报警通知邮件地址，可以设置多个
root@localhost
}
notification_email_from admin@stu31.com #设置邮件的发送地址
smtp_server 127.0.0.1 #设置 smtp server 的地址, 该地址必须是存在的
smtp_connect_timeout 30 #设置连接 smtp server 的超时时间
router_id LVS_DEVEL #运行 Keepalived 服务器的标识，发邮件时显示在邮件标题中的信息
}
vrrp_instance VI_1 {#定义 VRRP 实例, 实例名自定义
state MASTER #指定 Keepalived 的角色，MASTER 为主服务器，BACKUP 为备用服务器
interface eth0 #指定 HA 监测的接口
virtual_router_id 51 #虚拟路由标识，这个标识是一个数字(1-255)，在一个 VRRP 实例中主备服务器 ID 必须一样
priority 100 #优先级，数字越大优先级越高，在一个实例中主服务器优先级要高于备服务器
advert_int 1 #设置主备之间同步检查的时间间隔单位秒
authentication {#设置验证类型和密码
auth_type PASS #验证类型有两种{PASS|HA}
auth_pass Oracle #设置验证密码，在一个实例中主备密码保持一样
}
virtual_ipaddress {#定义虚拟 IP 地址, 可以有多个，每行一个
172.16.31.188
}
}
virtual_server 172.16.31.188 {#设置虚拟服务器，需要指定虚拟 IP 与服务端口, 用空格分隔
delay_loop 6 #设置健康状态检查时间，单位为秒
lb_algo rr #设置负载高度算法，rr 为轮询
lb_kind DR #设置 LVS 实现负载均衡的机制，可以为{NAT|TUN|DR} 三种
nat_mask 255.255.0.0 #设置掩码
persistence_timeout 50 #会话保持时间，单位为秒; 这个选项对于动态网页是非常有用的，为集群系统中 session 共享提供了一个很好的解决方案
protocol TCP #指定转发协议类型可以设置 {TCP|UDP} 两种
real_server 172.16.31.50 80 {#后端服务器节点, 需要指定 Real_server 的 IP 与端口, 用空格分隔
weight 1 #配置服务节点的权重, 数字越大，权重越高
HTTP_GET {#设置检测 Realserver 的方式为 http 协议
url {
path /
status_code 200 #设定返回状态码为 200 表示 Realserver 是存活的
}
connect_timeout 3 #设置响应超时时间
nb_get_retry 3 #设置超时重试次数
delay_before_retry 3 #设置超时后重试间隔
}
}
real_server 172.16.31.51 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
}

3. 将 Master 服务器 proxy 节点上的主配置文件拷贝到 Backup 服务器 proxy2 节点稍作修改
# scp /etc/keepalived/keepalived.conf proxy2:/etc/keepalived/
###### 修改如下两项
[root@proxy2 ~]# vim /etc/keepalived/keepalived.conf
state BACKUP
priority 99

4. 启动两台服务器上的 Keepalived 服务并设置为开机自启动
# chkconfig –add keepalived ; ssh proxy2 “chkconfig –add keepalived”
# chkconfig keepalived on ; ssh proxy2 “chkconfig keepalived on”
# chkconfig –list keepalived; ssh proxy2 “chkconfig –list keepalived”
keepalived 0:off 1:off 2:on 3:on 4:on 5:on 6:off
keepalived 0:off 1:off 2:on 3:on 4:on 5:on 6:off

启动服务：
# service keepalived start; ssh proxy2 “service keepalived start”

5. 观察各节点的服务启动后的日志：
MASTER 节点的日志；
[root@proxy keepalived]# tail -f /var/log/messages
Jan 17 15:57:42 proxy Keepalived[15019]: Starting Keepalived v1.2.13 (10/15,2014)
Jan 17 15:57:42 proxy Keepalived[15021]: Starting Healthcheck child process, pid=15024
Jan 17 15:57:42 proxy Keepalived[15021]: Starting VRRP child process, pid=15025
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Netlink reflector reports IP 172.16.31.52 added
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Netlink reflector reports IP fe80::a00:27ff:fe3b:2360 added
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Registering Kernel netlink reflector
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Registering Kernel netlink command channel
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Registering gratuitous ARP shared channel
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Netlink reflector reports IP 172.16.31.52 added
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Netlink reflector reports IP fe80::a00:27ff:fe3b:2360 added
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Registering Kernel netlink reflector
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Registering Kernel netlink command channel
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Opening file ‘/etc/keepalived/keepalived.conf’.
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Configuration is using : 63202 Bytes
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: Using LinkWatch kernel netlink reflector…
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Opening file ‘/etc/keepalived/keepalived.conf’.
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Configuration is using : 16989 Bytes
Jan 17 15:57:42 proxy Keepalived_vrrp[15025]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Using LinkWatch kernel netlink reflector…
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Activating healthchecker for service [172.16.31.50]:8080
Jan 17 15:57:42 proxy Keepalived_healthcheckers[15024]: Activating healthchecker for service [172.16.31.51]:8080
Jan 17 15:57:43 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) Transition to MASTER STATE
Jan 17 15:57:44 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) Entering MASTER STATE
Jan 17 15:57:44 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) setting protocol VIPs.
Jan 17 15:57:44 proxy Keepalived_healthcheckers[15024]: Netlink reflector reports IP 172.16.31.188 added
Jan 17 15:57:44 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 172.16.31.188

BACKUP 节点的日志：
[root@proxy2 keepalived]# tail -f /var/log/messages
Jan 17 15:57:42 proxy2 Keepalived[15005]: Starting Keepalived v1.2.13 (10/15,2014)
Jan 17 15:57:42 proxy2 Keepalived[15007]: Starting Healthcheck child process, pid=15009
Jan 17 15:57:42 proxy2 Keepalived[15007]: Starting VRRP child process, pid=15010
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Netlink reflector reports IP 172.16.31.53 added
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Netlink reflector reports IP 172.16.31.53 added
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Netlink reflector reports IP fe80::a00:27ff:fe6e:bd28 added
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Registering Kernel netlink reflector
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Registering Kernel netlink command channel
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Registering gratuitous ARP shared channel
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Netlink reflector reports IP fe80::a00:27ff:fe6e:bd28 added
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Registering Kernel netlink reflector
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Registering Kernel netlink command channel
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Opening file ‘/etc/keepalived/keepalived.conf’.
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Opening file ‘/etc/keepalived/keepalived.conf’.
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Configuration is using : 16991 Bytes
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Configuration is using : 63204 Bytes
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: Using LinkWatch kernel netlink reflector…
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) Entering BACKUP STATE
Jan 17 15:57:42 proxy2 Keepalived_vrrp[15010]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Using LinkWatch kernel netlink reflector…
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Activating healthchecker for service [172.16.31.50]:8080
Jan 17 15:57:42 proxy2 Keepalived_healthcheckers[15009]: Activating healthchecker for service [172.16.31.51]:8080

6. 开启 Master 节点与 Backup 节点服务器的路由转发功能
#sed -i ‘s/net.ipv4.ip_forward = 0/net.ipv4.ip_forward = 1/g’ /etc/sysctl.conf

Linux 服务器 LB 群集之 LVS-NAT http://www.linuxidc.com/Linux/2013-05/84774.htm

Linux 下群集服务之 LB 集群 -LVS-NAT 模式 http://www.linuxidc.com/Linux/2012-05/59839.htm

LVS-NAT+ipvsadm 实现 RHEL 5.7 上的服务集群 http://www.linuxidc.com/Linux/2012-04/58936.htm

Linux 虚拟服务器 LVS-NAT 和 LVS-DR 模型的实现 http://www.linuxidc.com/Linux/2011-09/42934.htm

更多详情见请继续阅读下一页的精彩内容：http://www.linuxidc.com/Linux/2015-01/112560p2.htm

二. 安装配置后端的 tomcat 服务器节点
1. 配置后端 tomcat 服务器节点；
具体配置参考我的前篇博客，都是连续性的实验；
配置完成后，访问测试如图：

tom1 节点：
LVS+Keepalived 高可用负载均衡集群架构实验

tom2 节点：
LVS+Keepalived 高可用负载均衡集群架构实验

实现了负载均衡访问；

2. 为两台 Realserver 提供 Sysv 格式的脚本来自动修改内核参数与虚拟 IP 并运行脚本
#vim rs.sh
. /etc/rc.d/init.d/functions
VIP=172.16.31.188
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

注释：脚本中的 VIP 定义的是虚拟 IP 地址

3. 脚本两个节点都存放一份；然后执行此脚本：
#sh rs.sh start

4. 查看设置后的状态：
12 # sh rs.sh status
LVS-DR real server Running.

查看节点的 VIP 地址设置完成与否：
tom1 节点；
[root@tom1 ~]# ip addr show lo
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
inet 172.16.31.188/32 brd 172.16.31.188 scope global lo:0
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever

tom2 节点：
[root@tom2 ~]# ip addr show lo
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
inet 172.16.31.188/32 brd 172.16.31.188 scope global lo:0
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever

5. 查看 MASTER 节点 proxy 的 ipvs 规则：
1234567 [root@proxy keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.16.31.188:0 rr persistent 50
-> 172.16.31.50:8080 Route 1 0 0
-> 172.16.31.51:8080 Route 1 0 0

从规则中可以看出虚拟 IP 与 Port 及调度算法为 rr；其中有两个 Realserver

三. 访问测试及故障模拟 ces
1. 访问测试：
LVS+Keepalived 高可用负载均衡集群架构实验

由上可知，使用的是 rr 调度算法，在访问测试时可能需要多访问几次或换个浏览器来测试访问

2. 模拟 Master 服务器出现故障，将 Master 主机上的 Keepalived 服务停止，查看 Backup 服务器是否接管所有服务
[root@proxy keepalived]# service keepalived stop
Stopping keepalived: [OK]

3. 观察节点切换后的系统日志：
MASTER 节点的日志：
[root@proxy keepalived]# tail -f /var/log/messages
Jan 17 16:20:40 proxy Keepalived[15021]: Stopping Keepalived v1.2.13 (10/15,2014)
Jan 17 16:20:40 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) sending 0 priority
Jan 17 16:20:40 proxy Keepalived_vrrp[15025]: VRRP_Instance(VI_1) removing protocol VIPs.
Jan 17 16:20:40 proxy Keepalived_healthcheckers[15024]: Netlink reflector reports IP 172.16.31.188 removed
Jan 17 16:20:40 proxy Keepalived_healthcheckers[15024]: Removing service [172.16.31.50]:8080 from VS [172.16.31.188]:0
Jan 17 16:20:40 proxy Keepalived_healthcheckers[15024]: Removing service [172.16.31.51]:8080 from VS [172.16.31.188]:0
Jan 17 16:20:40 proxy kernel: IPVS: __ip_vs_del_service: enter

BACKUP 节点的日志：
1234567 [root@proxy2 keepalived]# tail -f /var/log/messages
Jan 17 16:20:41 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) Transition to MASTER STATE
Jan 17 16:20:42 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) Entering MASTER STATE
Jan 17 16:20:42 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) setting protocol VIPs.
Jan 17 16:20:42 proxy2 Keepalived_healthcheckers[15009]: Netlink reflector reports IP 172.16.31.188 added
Jan 17 16:20:42 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 172.16.31.188
Jan 17 16:20:47 proxy2 Keepalived_vrrp[15010]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 172.16.31.188

4. 查看各节点的 VIP 地址状态及 IPVS 规则：
原 MASTER 节点：
[root@proxy keepalived]# ip addr show
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 08:00:27:3b:23:60 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.52/16 brd 172.16.255.255 scope global eth0
inet6 fe80::a00:27ff:fe3b:2360/64 scope link
valid_lft forever preferred_lft forever

1234 [root@proxy keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn

由上可见 Master 服务器上已删除虚拟 IP 与 LVS 规则

BACKUP 节点：
[root@proxy2 keepalived]# ip addr show
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 08:00:27:6e:bd:28 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.53/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.188/32 scope global eth0
inet6 fe80::a00:27ff:fe6e:bd28/64 scope link
valid_lft forever preferred_lft forever

由上可见，虚拟 IP 地址已成功在 Backup 服务器启动
1234567 [root@proxy2 keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.16.31.188:0 rr persistent 50
-> 172.16.31.50:8080 Route 1 0 0
-> 172.16.31.51:8080 Route 1 0 0

LVS 的规则也已成功配置在 Backup 服务器上面

5. 再次访问测试服务器是否正常提供服务
LVS+Keepalived 高可用负载均衡集群架构实验

6. 假如 Master 服务器修复好已重新上线，则虚拟 IP 地址与 LVS 规则会重新配置到 Master 服务器上而在 Backup 服务器上删除
[root@proxy keepalived]# service keepalived start
Starting keepalived: [OK]

观察启动日志：
MASTER 节点的日志：
[root@proxy keepalived]# tail -f /var/log/messages
Jan 17 16:34:50 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) Transition to MASTER STATE
Jan 17 16:34:50 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) Received lower prio advert, forcing new election
Jan 17 16:34:51 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) Entering MASTER STATE
Jan 17 16:34:51 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) setting protocol VIPs.
Jan 17 16:34:51 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 172.16.31.188
Jan 17 16:34:51 proxy Keepalived_healthcheckers[15365]: Netlink reflector reports IP 172.16.31.188 added
Jan 17 16:34:56 proxy Keepalived_vrrp[15366]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 172.16.31.188

自动进入了 MASTER 状态；

BACKUP 节点的日志：
12345 [root@proxy2 keepalived]# tail -f /var/log/messages
Jan 17 16:34:50 proxy2 Keepalived_vrrp[15258]: VRRP_Instance(VI_1) Received higher prio advert
Jan 17 16:34:50 proxy2 Keepalived_vrrp[15258]: VRRP_Instance(VI_1) Entering BACKUP STATE
Jan 17 16:34:50 proxy2 Keepalived_vrrp[15258]: VRRP_Instance(VI_1) removing protocol VIPs.
Jan 17 16:34:50 proxy2 Keepalived_healthcheckers[15257]: Netlink reflector reports IP 172.16.31.188 removed

自动进入 BACKUP 状态，移除 VIP；

查看各节点的 VIP 和 IPVS 规则：
MASTER 节点的的 VIP 生成成功，IPVS 规则生成：
[root@proxy keepalived]# ip addr show
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 08:00:27:3b:23:60 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.52/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.188/32 scope global eth0
inet6 fe80::a00:27ff:fe3b:2360/64 scope link
valid_lft forever preferred_lft forever
[root@proxy keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.16.31.188:0 rr persistent 50
-> 172.16.31.50:8080 Route 1 0 0
-> 172.16.31.51:8080 Route 1 0 0

BACKUP 节点自动降级，VIP 自动移除，但是 ipvs 规则还是存在的，这对我们是没有影响的，没有了 IP 地址只有规则也是不生效的
[root@proxy2 keepalived]# ip addr show
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 08:00:27:6e:bd:28 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.53/16 brd 172.16.255.255 scope global eth0
inet6 fe80::a00:27ff:fe6e:bd28/64 scope link
valid_lft forever preferred_lft forever
[root@proxy2 keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.16.31.188:0 rr persistent 50
-> 172.16.31.50:8080 Route 1 0 0
-> 172.16.31.51:8080 Route 1 0 0

7. 如果后端 Realserver 出现故障，则 LVS 规则会清除相应 Realserver 的规则
我们将后端的 tom1 节点的 tomcat 服务停止：
[root@tom1 ~]# catalina.sh stop
Using CATALINA_BASE: /usr/local/tomcat
Using CATALINA_HOME: /usr/local/tomcat
Using CATALINA_TMPDIR: /usr/local/tomcat/temp
Using JRE_HOME: /usr/Java/latest
Using CLASSPATH: /usr/local/tomcat/bin/bootstrap.jar:/usr/local/tomcat/bin/tomcat-juli.jar
[root@tom1 ~]#

观察 LVS 节点的启动日志：

MASTER 节点：
出现连接错误，发送了邮件：
[root@proxy keepalived]# tail -f /var/log/messages
Jan 17 16:40:39 proxy Keepalived_healthcheckers[15365]: Error connecting server [172.16.31.50]:8080.
Jan 17 16:40:39 proxy Keepalived_healthcheckers[15365]: Removing service [172.16.31.50]:8080 from VS [172.16.31.188]:0
Jan 17 16:40:39 proxy Keepalived_healthcheckers[15365]: Remote SMTP server [127.0.0.1]:25 connected.
Jan 17 16:40:39 proxy Keepalived_healthcheckers[15365]: SMTP alert successfully sent.

BACKUP 节点：
出现连接错误，发送了邮件：
[root@proxy2 keepalived]# tail -f /var/log/messages
Jan 17 16:40:42 proxy2 Keepalived_healthcheckers[15257]: Error connecting server [172.16.31.50]:8080.
Jan 17 16:40:42 proxy2 Keepalived_healthcheckers[15257]: Removing service [172.16.31.50]:8080 from VS [172.16.31.188]:0
Jan 17 16:40:42 proxy2 Keepalived_healthcheckers[15257]: Remote SMTP server [127.0.0.1]:25 connected.
Jan 17 16:40:42 proxy2 Keepalived_healthcheckers[15257]: SMTP alert successfully sent.

在到 MASTER 节点查看 ipvs 规则：
1234567 [root@proxy keepalived]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.16.31.188:0 rr persistent 50
-> 172.16.31.51:8080 Route 1 0 0
You have new mail in /var/spool/mail/root

规则已经将故障的后端 tomcat 服务器移除，并且我们还收到了邮件；查看邮件：
[root@proxy keepalived]# mail
Heirloom Mail version 12.4 7/29/08. Type ? for help.
“/var/spool/mail/root”: 16 messages 12 new 14 unread
略 …..
N 12 admin@stu31.com Sat Jan 17 15:44 17/594 “[LVS_DEVEL] Realserver [172.16.31.51]:80 – DOWN”
N 13 admin@stu31.com Sat Jan 17 15:44 17/594 “[LVS_DEVEL] Realserver [172.16.31.50]:80 – DOWN”
N 14 admin@stu31.com Sat Jan 17 15:50 17/594 “[LVS_DEVEL] Realserver [172.16.31.51]:80 – DOWN”
N 15 admin@stu31.com Sat Jan 17 15:50 17/594 “[LVS_DEVEL] Realserver [172.16.31.50]:80 – DOWN”
N 16 admin@stu31.com Sat Jan 17 16:40 17/596 “[LVS_DEVEL] Realserver [172.16.31.50]:8080 – DOWN”
& 16
Message 16:
From admin@stu31.com Sat Jan 17 16:40:39 2015
Return-Path: <admin@stu31.com>
X-Original-To: root@localhost
Delivered-To: root@localhost.stu31.com
Date: Sat, 17 Jan 2015 08:40:39 +0000
From: admin@stu31.com
Subject: [LVS_DEVEL] Realserver [172.16.31.50]:8080 – DOWN
X-Mailer: Keepalived
To: root@localhost.stu31.com
Status: R
=> CHECK failed on service : connection error <=
& quit
Held 16 messages in /var/spool/mail/root

注：我在这里还测试了优先级相同时，LVS 的主备节点的选举情况，如果优先级都是 100 的话，我们停止 MASTER 节点的 keepalived 服务后，VIP 和 IPVS 规则自动转移到了 BACKUP 节点，但是我们将 MASTER 节点的 keepalived 服务启动后，VIP 还是在 BACKUP 节点上，ipvs 规则会在 MASTER 节点上生成；
所以我们在生产环境中一定要注意优先级的设置，主节点的优先级必须比备用节点的优先级高；

至此，LVS+Keepalived 实现高可用负载均衡集群已全部完成

CentOS 6.3 下 Haproxy+Keepalived+Apache 配置笔记 http://www.linuxidc.com/Linux/2013-06/85598.htm

Haproxy + KeepAlived 实现 WEB 群集 on CentOS 6 http://www.linuxidc.com/Linux/2012-03/55672.htm

Keepalived+Haproxy 配置高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/56748.htm

Haproxy+Keepalived 构建高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/55880.htm

CentOS 7 上配置 LVS + Keepalived + ipvsadm http://www.linuxidc.com/Linux/2014-11/109237.htm

Keepalived 高可用集群搭建 http://www.linuxidc.com/Linux/2014-09/106965.htm

Keepalived 的详细介绍：请点这里
Keepalived 的下载地址：请点这里