keepalived是一款用C编写的，旨在给linux系统和基于linux的设施提供简单、稳固的高可用和负载均衡功能的软件。它基于linux内核的ipvs模块实现4层负载均衡，基于VRRP协议实现服务的高可用。
一、VRRP协议
VRRP（Virtual Router Redundancy Protocol，虚拟路由冗余协议）是一种容错协议。通常，一个网络内的所有主机都设置一条默认路由，这样，主机发出的目的地址不在本网段的报文将被通过默认路由发往路由器RouterA，从而实现了主机与外部网络的通信。当路由器RouterA 坏掉时，本网段内所有以RouterA 为默认路由下一跳的主机将无法与外部通信，这就是单点故障。VRRP就是为解决上述问题而提出的。
VRRP 将局域网的一组路由器组织成一个虚拟路由器，称之为一个备份组。这个虚拟的路由器拥有自己的IP地址（vip，也就是路由器所在局域网内其他机器的默认路由）和MAC地址（vmac），占有这个IP的物理路由器作为master实际负责ARP响应和数据包转发，组中的其它路由器作为备份（backup）的角色处于待命状态。master会发组播消息，当backup在超时时间内收不到vrrp包时就认为master宕掉了，这时就需要根据VRRP的优先级来选举一个backup充当新的master，继续向网络内的主机提供路由服务。从而实现网络内的主机不间断地与外部网络进行通信。
每个Router都有一个 1-255 之间的优先级，级别最高的（highest priority）将成为master路由器，若优先级相同，则IP地址较大者胜出。通过降低master的优先级可以让处于backup状态的路由器抢占主路由器的状态，接管虚拟IP。
二、keepalived的组件
keepalived主要有三大模块：
core：keepalived的核心，负责主进程的启动和维护，全局配置文件的加载解析等；
check：负责健康状态检测，包括各种常见的健康检查方式；
vrrp：实现VRRP协议；
三、keepalive的适用场景
keepalived理论上可以为mysqld，httpd等服务提供高可用，给这些服务做高可用需要配置共享存储资源，对此keepalived需要借助额外的命令或脚本实现，这种情况下keepalived的性能是显然不如heartbeat或corosync的。keepalived适合提供轻量级的高可用方案，如对作为反向代理的nginx、haproxy以及ipvs做高可用，给这些服务做高可用都无需配置共享存储资源。对ipvs做高可用时，能运用多种检测机制获知后端real server的健康状态，动态和自适应地维护、管理服务器池。
四、使用keepalived做反向代理nginx的高可用
1、实验拓扑图
2、在node3上安装httpd，在node1和node2上安装nginx并配置为反向代理服务器，代理至node3；另外，要确保做高可用的两个节点时间同步。
[root@node3 ~]# yum -y install httpd ... [root@node3 ~]# vim /var/www/html/test.html #创建一个测试页面 hello,keepalived [root@node3 ~]# service httpd start ...
[root@node1 ~]# yum -y install nginx;ssh root@node2 ‘yum -y install nginx‘
...
[root@node1 ~]# vim /etc/nginx/conf.d/default.conf
...
location ~ \.html$ {
proxy_pass http://192.168.30.13;
}
...
[root@node1 ~]# scp /etc/nginx/conf.d/default.conf root@node2:/etc/nginx/conf.d/
...3、在node1和node2上安装配置keepalived
yum -y install keepalived
主配置文件：/etc/keepalived/keepalived.conf
查看配置文件帮助信息：man keepalived.conf
keepalived的日志信息会输出至/var/log/messages
[root@node1 ~]# yum -y install keepalived;ssh root@node2 ‘yum -y install keepalived‘
...
[root@node1 ~]# rpm -ql keepalived
/etc/keepalived
/etc/keepalived/keepalived.conf #配置文件
/etc/rc.d/init.d/keepalived
/etc/sysconfig/keepalived
/usr/bin/genhash
/usr/libexec/keepalived
/usr/sbin/keepalived
...
/usr/share/doc/keepalived-1.2.13/samples #该目录下有一些样例文件，供参考
...
[root@node1 ~]# ls /usr/share/doc/keepalived-1.2.13/samples
keepalived.conf.fwmark keepalived.conf.misc_check_arg keepalived.conf.status_code keepalived.conf.vrrp.localcheck keepalived.conf.vrrp.sync
keepalived.conf.HTTP_GET.port keepalived.conf.quorum keepalived.conf.track_interface keepalived.conf.vrrp.lvs_syncd sample.misccheck.smbcheck.sh
keepalived.conf.inhibit keepalived.conf.sample keepalived.conf.virtualhost keepalived.conf.vrrp.routes
keepalived.conf.IPv6 keepalived.conf.SMTP_CHECK keepalived.conf.virtual_server_group keepalived.conf.vrrp.scripts
keepalived.conf.misc_check keepalived.conf.SSL_GET keepalived.conf.vrrp keepalived.conf.vrrp.static_ipaddress
[root@node1 ~]# cd /etc/keepalived/
[root@node1 keepalived]# mv keepalived.conf keepalived.conf.back
[root@node1 keepalived]# vim keepalived.conf
! Configuration File for keepalived #keepalived配置文件中，以“!”开头的为注释信息
global_defs { #该区域主要是配置故障发生时的通知对象以及机器标识
notification_email { #指定keepalived在发生事件（如切换、故障）时发送email给谁，多个写多行
root@localhost #收件人
magedu@126.com
}
notification_email_from kanotify@magedu.com #发件人
smtp_connect_timeout 10 #连接smtp服务器的超时时长
smtp_server 127.0.0.1 #smtp服务器地址
router_id nginx-node1 #机器标识；通常为hostname，但不是必须为hostname
}
vrrp_script chk_nginx { #健康状态检测
script "killall -0 nginx" #信号0用来判断进程是否存在；脚本返回非0表示失败
interval 1 #检测的间隔时长
weight -2 #确定失败则将优先级减2
fall 3 #确定为失败需要检测的次数
rise 1 #确定为正常需要检测的次数
}
# 说明：上面配置的健康状态检测机制是有缺陷的，它仅能获知服务的运行状态，但实际上当服务异常时并不能实现资源正常、合理地转移。此段不用配置
vrrp_script chk_mantaince_down {
script "[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
# 通过手动创建文件的方式使节点降级
interval 1
weight -2
fall 3
rise 1
}
vrrp_instance VI_1 { #定义一个vrrp实例，可定义多个，每个vrrp实例名称要惟一
interface eth0 #将虚拟ip绑定在哪个接口上
state MASTER #指定节点的初始状态；但这里指定的不算，keepalived启动时根据节点的优先级高低来确定节点的角色
priority 100 #优点级，1-255
virtual_router_id 11 #虚拟路由id，用来区分多个instance的VRRP组播
garp_master_delay 1 #当切为master状态后多久更新ARP缓存，默认5秒
authentication { #认证区域，认证类型有PASS和HA（IPSEC）
auth_type PASS
auth_pass magedu
}
track_interface { #跟踪接口的状态
eth0
}
virtual_ipaddress {
192.168.30.30/24 dev eth0 label eth0:0 #虚拟IP地址，会添加在master上
}
track_script { #引用vrrp脚本，即在 vrrp_script 部分指定的名称
chk_nginx
chk_mantaince_down
}
notify_master "/etc/keepalived/notify.sh master"
notify_backup "/etc/keepalived/notify.sh backup"
notify_fault "/etc/keepalived/notify.sh fault"
# notify_master/backup/fault 分别表示节点状态为主/备/出错时所执行的脚本
}
# 其它常用参数：
nopreempt 表示不抢占。允许一个priority较低的节点保持master状态，即使priority更高的节点恢复正常
要配置非抢占模式，在优先级较高节点的配置文件中：
state BACKUP
nopreempt
use_vmac 是否使用VRRP的虚拟MAC地址
notify 表示任何一状态切换时都会调用该脚本
smtp_alert 表示开启邮件通知（用全局区域的邮件设置来发通知）
[root@node1 keepalived]# vim notify.sh #创建一个通知脚本
#!/bin/bash
# Author: MageEdu <linuxedu@foxmail.com>
# description: An example of notify script
#
vip=192.168.30.30
contact=‘root@localhost‘
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
}
case "$1" in
master)
notify master
/etc/rc.d/init.d/nginx start
exit 0
;;
backup)
notify backup
/etc/rc.d/init.d/nginx stop
exit 0
;;
fault)
notify fault
/etc/rc.d/init.d/nginx stop
exit 0
;;
*)
echo ‘Usage: `basename $0` {master|backup|fault}‘
exit 1
;;
esac
[root@node1 keepalived]# chmod +x notify.sh
[root@node1 keepalived]# scp keepalived.conf notify.sh root@node2:/etc/keepalived/[root@node2 ~]# ls /etc/keepalived/ keepalived.conf keepalived.conf.back notify.sh [root@node2 ~]# vim /etc/keepalived/keepalived.conf #对node2上的keepalived配置文件做适当修改 ... router_id nginx-node2 ... state BACKUP priority 99 #从节点优点级较低 ... [root@node2 ~]# service keepalived start #先尝试启动从节点 Starting keepalived: [ OK ] [root@node2 ~]# tail -f /var/log/messages ... May 26 21:33:52 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Entering BACKUP STATE #刚开始按照配置文件中的设定进入backup状态 May 26 21:33:52 node2 Keepalived_vrrp[46904]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)] May 26 21:33:52 node2 Keepalived_vrrp[46904]: VRRP_Script(chk_mantaince_down) succeeded May 26 21:33:56 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Transition to MASTER STATE #因node1上的keepalived还没启动，所以node2会转换为master状态 May 26 21:33:57 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Entering MASTER STATE May 26 21:33:57 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) setting protocol VIPs. May 26 21:33:57 node2 Keepalived_healthcheckers[46902]: Netlink reflector reports IP 192.168.30.30 added May 26 21:33:57 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.30.30 May 26 21:33:57 node2 Keepalived_vrrp[46904]: VRRP_Script(chk_nginx) succeeded May 26 21:33:58 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.30.30 ... [root@node2 keepalived]# ip addr show #可以看到vip已配置 ... 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:bd:68:23 brd ff:ff:ff:ff:ff:ff inet 192.168.30.20/24 brd 192.168.30.255 scope global eth0 inet 192.168.30.30/24 scope global secondary eth0:0 inet6 fe80::20c:29ff:febd:6823/64 scope link valid_lft forever preferred_lft forever
[root@node1 ~]# service keepalived start Starting keepalived: [ OK ] [root@node1 ~]# tail -f /var/log/messages ... May 26 21:34:37 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) Transition to MASTER STATE May 26 21:34:37 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) Received lower prio advert, forcing new election #收到较低的优先级通告，强制进行新的选举 May 26 21:34:38 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) Entering MASTER STATE #进入master状态 May 26 21:34:38 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) setting protocol VIPs. #设置vip May 26 21:34:38 node1 Keepalived_healthcheckers[43126]: Netlink reflector reports IP 192.168.30.30 added May 26 21:34:38 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.30.30 May 26 21:34:39 node1 Keepalived_vrrp[43127]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.30.30 [root@node1 ~]# ip addr show #可以看到node1已抢到vip 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:29:40:35:9d brd ff:ff:ff:ff:ff:ff inet 192.168.30.10/24 brd 192.168.30.255 scope global eth0 inet 192.168.30.30/24 scope global secondary eth0:0 inet6 fe80::20c:29ff:fe40:359d/64 scope link valid_lft forever preferred_lft forever [root@node1 ~]# service nginx status #nginx已正常启动 nginx (pid 53134) is running... [root@node1 ~]# mail #已收到角色切换的消息 Heirloom Mail version 12.4 7/29/08. Type ? for help. "/var/spool/mail/root": 1 message 1 new >N 1 root Thu May 26 21:35 20/732 "node1 to be master: 192.168.30.30 floating"
[root@node2 ~]# tail -f /var/log/messages #node2上配置的vip已被移除 ... May 26 21:34:37 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Received higher prio advert May 26 21:34:37 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) Entering BACKUP STATE May 26 21:34:37 node2 Keepalived_vrrp[46904]: VRRP_Instance(VI_1) removing protocol VIPs. May 26 21:34:37 node2 Keepalived_healthcheckers[46902]: Netlink reflector reports IP 192.168.30.30 removed May 26 21:34:39 node2 Keepalived_vrrp[46904]: VRRP_Script(chk_nginx) failed
4、测试
[root@node5 ~]# curl 192.168.30.30/test.html hello,keepalived
模拟资源转移：
[root@node1 keepalived]# service keepalived stop Stopping keepalived: [ OK ] [root@node1 keepalived]# ip addr show ... 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:40:35:9d brd ff:ff:ff:ff:ff:ff inet 192.168.30.10/24 brd 192.168.30.255 scope global eth0 inet6 fe80::20c:29ff:fe40:359d/64 scope link valid_lft forever preferred_lft forever [root@node1 keepalived]# service nginx status nginx is stopped
[root@node2 keepalived]# ip addr show ... 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:bd:68:23 brd ff:ff:ff:ff:ff:ff inet 192.168.30.20/24 brd 192.168.30.255 scope global eth0 inet 192.168.30.30/24 scope global secondary eth0:0 #vip已转到node2上 inet6 fe80::20c:29ff:febd:6823/64 scope link valid_lft forever preferred_lft forever You have new mail in /var/spool/mail/root [root@node2 keepalived]# service nginx status nginx (pid 11412) is running...
[root@node5 ~]# curl 192.168.30.30/test.html hello,keepalived
五、使用keepalived做ipvs的高可用，采用双主模型
1、实验拓扑图
2、配置好两个后端real server
[root@node3 ~]# echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore [root@node3 ~]# echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore [root@node3 ~]# echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce [root@node3 ~]# echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce [root@node3 ~]# ifconfig lo:0 192.168.30.31 netmask 255.255.255.255 broadcast 192.168.30.31 up [root@node3 ~]# ifconfig lo:1 192.168.30.32 netmask 255.255.255.255 broadcast 192.168.30.32 up [root@node3 ~]# route add -host 192.168.30.31 dev lo:0 [root@node3 ~]# route add -host 192.168.30.32 dev lo:1 [root@node3 ~]# cd /var/www/html [root@node3 html]# ls [root@node3 html]# vim index.html hello [root@node3 html]# vim test.html hello,this is node3 [root@node3 html]# service httpd start ... # 在node4上执行类似的步骤
3、对ipvs做双主模型的高可用，并将两个ipvs节点作为后端RS出现故障时的备用服务器
要配置双主模型，需要配置两个vrrp实例，两个节点在一个实例中的角色为分别为主/从，在另一个实例中则刚好反过来，为从/主
[root@node1 ~]# yum -y install httpd;ssh root@node2 ‘yum -y install httpd‘
...
[root@node1 ~]# vim /var/www/html/test.html
fallback1
[root@node1 ~]# service httpd start
...
[root@node1 ~]# cd /etc/keepalived/
[root@node1 keepalived]# vim keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
root@localhost
magedu@126.com
}
notification_email_from kanotify@magedu.com
smtp_connect_timeout 10
smtp_server 127.0.0.1
router_id LVS_DEVEL
}
vrrp_script chk_schedown1 {
script "[[ -f /etc/keepalived/down1 ]] && exit 1 || exit 0"
interval 2
weight -2
}
vrrp_script chk_schedown2 {
script "[[ -f /etc/keepalived/down2 ]] && exit 1 || exit 0"
interval 2
weight -2
}
vrrp_instance VI_1 { #第一个vrrp实例
interface eth0
state MASTER #将node1配置为第一个实例中的master
priority 100
virtual_router_id 51
garp_master_delay 1
authentication {
auth_type PASS
auth_pass magedu
}
track_interface {
eth0
}
virtual_ipaddress {
192.168.30.31/24 dev eth0 label eth0:0 #vip1
}
track_script {
chk_schedown1
}
}
virtual_server 192.168.30.31 80 { #定义一个集群服务
delay_loop 6
lb_algo rr #lvs调度算法
lb_kind DR #lvs模型
persistence_timeout 30 #持久连接
protocol TCP
sorry_server 127.0.0.1 80 #定义后端real server出现故障时的备用服务器
real_server 192.168.30.13 80 { #定义后端RS
weight 1
HTTP_GET { #定义后端RS健康状态检测的方式
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
# RS健康状态检测方式有多种：HTTP_GET|SSL_GET|TCP_CHECK|SMTP_CHECK|MISC_CHECK，具体用法可man keepalived.conf
# 如果要使用TCP_CHECK检测各real server的健康状态，那么，上面HTTP_GET部分的定义可以替换为如下内容：
# TCP_CHECK {
# connect_port 80
# connect_timeout 3
# }
real_server 192.168.30.14 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
}
vrrp_instance VI_2 { #第二个vrrp实例
interface eth0
state BACKUP
priority 99
virtual_router_id 52 #每个vrrp实例的虚拟路由ID必须惟一
garp_master_delay 1
authentication {
auth_type PASS
auth_pass magedu
}
track_interface {
eth0
}
virtual_ipaddress {
192.168.30.32/24 dev eth0 label eth0:1 #vip2
}
track_script {
chk_schedown2
}
}
virtual_server 192.168.30.32 80 {
delay_loop 6
lb_algo rr
lb_kind DR
persistence_timeout 30
protocol TCP
sorry_server 127.0.0.1 80
real_server 192.168.30.13 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
real_server 192.168.30.14 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
}
[root@node1 keepalived]# scp keepalived.conf root@node2:/etc/keepalived/
keepalived.conf 100% 2694 2.6KB/s 00:00[root@node2 ~]# vim /var/www/html/test.html
fallback2
[root@node2 ~]# service httpd start
...
[root@node2 ~]# cd /etc/keepalived/
[root@node2 keepalived]# vim keepalived.conf #对node2上的keepalived配置文件做适当修改
...
vrrp_instance VI_1 {
...
state BACKUP
priority 99
...
}
...
vrrp_instance VI_2 {
...
state MASTER
priority 100
...
}
[root@node2 keepalived]# service keepalived start;ssh root@node1 ‘service keepalived start‘
Starting keepalived: [ OK ]
Starting keepalived: [ OK ]
[root@node2 keepalived]# yum -y install ipvsadm;ssh root@node1 ‘yum -y install ipvsadm‘
...
#安装ipvsadm只是便于查看ipvs规则，非必须
[root@node2 keepalived]# ip addr show #因node2在第二个vrrp实例中优先级列高，故vip2已配置在node2上
...
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:bd:68:23 brd ff:ff:ff:ff:ff:ff
inet 192.168.30.20/24 brd 192.168.30.255 scope global eth0
inet 192.168.30.32/24 scope global secondary eth0:1
inet6 fe80::20c:29ff:febd:6823/64 scope link
valid_lft forever preferred_lft forever
[root@node2 keepalived]# ipvsadm -L -n #ipvs规则已生成
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 192.168.30.31:80 rr persistent 30
-> 192.168.30.13:80 Route 1 0 0
-> 192.168.30.14:80 Route 1 0 0
TCP 192.168.30.32:80 rr persistent 30
-> 192.168.30.13:80 Route 1 0 0
-> 192.168.30.14:80 Route 1 0 0[root@node1 keepalived]# ip addr show #vip1已配置在node1上 ... 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:40:35:9d brd ff:ff:ff:ff:ff:ff inet 192.168.30.10/24 brd 192.168.30.255 scope global eth0 inet 192.168.30.31/24 scope global secondary eth0:0 inet6 fe80::20c:29ff:fe40:359d/64 scope link valid_lft forever preferred_lft forever [root@node1 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 192.168.30.31:80 rr persistent 30 -> 192.168.30.13:80 Route 1 0 0 -> 192.168.30.14:80 Route 1 0 0 TCP 192.168.30.32:80 rr persistent 30 -> 192.168.30.13:80 Route 1 0 0 -> 192.168.30.14:80 Route 1 0 0
4、测试
[root@node5 ~]# curl 192.168.30.31/test.html #因为启用了持久连接，所以请求可能在一段时间内始终被定向至同一RS hello,this is node4 [root@node5 ~]# curl 192.168.30.31/test.html hello,this is node4 [root@node5 ~]# curl 192.168.30.31/test.html hello,this is node3 [root@node5 ~]# curl 192.168.30.32/test.html hello,this is node3 [root@node5 ~]# curl 192.168.30.32/test.html hello,this is node3
[root@node4 html]# service httpd stop #将一个RS上的服务停掉 Stopping httpd: [ OK ]
[root@node2 keepalived]# ipvsadm -L -n #keepalived探测到后端RS的异常状态，将故障RS移除 IP Virtual Server version 1.2.1 (size=4096) Prot LocalAddress:Port Scheduler Flags -> RemoteAddress:Port Forward Weight ActiveConn InActConn TCP 192.168.30.31:80 rr persistent 30 -> 192.168.30.13:80 Route 1 0 0 TCP 192.168.30.32:80 rr persistent 30 -> 192.168.30.13:80 Route 1 0 1
[root@node5 ~]# curl 192.168.30.31/test.html hello,this is node3 [root@node5 ~]# curl 192.168.30.31/test.html hello,this is node3 [root@node5 ~]# curl 192.168.30.32/test.html hello,this is node3
[root@node3 html]# service httpd stop #将另一个RS上的服务也停掉 Stopping httpd: [ OK ]
[root@node2 ~]# ipvsadm -L -n #可以看到sorry server已被添加进ipvs规则中 IP Virtual Server version 1.2.1 (size=4096) Prot LocalAddress:Port Scheduler Flags -> RemoteAddress:Port Forward Weight ActiveConn InActConn TCP 192.168.30.31:80 rr persistent 30 -> 127.0.0.1:80 Local 1 0 0 TCP 192.168.30.32:80 rr persistent 30 -> 127.0.0.1:80 Local 1 0 0
[root@node5 ~]# curl 192.168.30.31/test.html fallback1 [root@node5 ~]# curl 192.168.30.32/test.html fallback2
原文地址：http://9124573.blog.51cto.com/9114573/1783867
