一、架构

二、唠叨一会原理：

1、nginx

Nginx进程基于于Master+Slave(worker)多进程模型，自身具有非常稳定的子进程管理功能。在Master进程分配模式下，Master进程永远不进行业务处理，只是进行任务分发，

从而达到Master进程的存活高可靠性，Slave(worker)进程所有的业务信号都 由主进程发出，Slave(worker)进程所有的超时任务都会被Master中止，属于非阻塞式任务模型。

2、keepalived

Keepalived是Linux下面实现VRRP 备份路由的高可靠性运行件。基于Keepalived设计的服务模式能够真正做到主服务器和备份服务器故障时IP瞬间无缝交接,作用:

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

3、单点故障

Nginx有很强代理功能，但是一台nginx就形成了单点，现在使用keepalived来解决这个问题，keepalived的故障转移时间很短.

Nginx+keepalived双机实现nginx反向代理服务的高可用,一台nginx挂掉之后不影响应用也不影响内网访问外网.

4、此架构需要考虑的问题

1) Master没挂，则Master占有vip且nginx运行在Master上

2) Master挂了，则backup抢占vip且在backup上运行nginx服务

3) 如果master服务器上的nginx服务挂了，则vip资源转移到backup服务器上

4) 检测后端服务器的健康状态

5、叙述

Master和Backup两边都开启nginx服务，无论Master还是Backup，当其中的一个keepalived服务停止后，vip都会漂移到keepalived服务还在的节点上，

如果要想使nginx服务挂了，vip也漂移到另一个节点，则必须用脚本或者在配置文件里面用shell命令来控制。

首先必须明确后端服务器的健康状态检测keepalived在这种架构上是无法检测的，后端服务器的健康状态检测是有nginx来判断的，但是nginx的检测机制有一定的缺陷，后端服务器某一个宕机之后，nginx还是会分发请求给它，在一定的时间内后端服务响应不了，nginx则会发给另外一个服务器，然后当客户的请求来了，nginx会一段时间内不会把请求分发给已经宕机的服务器，但是过一段时间后，nginx还是会把分发请求发给宕机的服务器上。

三、准备工作

1、环境
系统版本：CentOS 6.4 64bit
nginx版本：nginx-1.4.7.tar.gz
keepalived版本：keepalived-1.2.7.tar.gz2、地址规划
nginx+keepalived主      192.168.0.189     master
nginx+keepalived辅      192.168.0.200     badkup
vip                 192.168.0.250     vip

四、安装nginx和keepalived服务

1、安装nginx服务[分别在master机器和backup机器安装nginx服务]

#！/bin/bash
# time 2014-08-18
# Allentuns#安装依赖环境包
yum -y install gcc gcc-c++ make unzip wget vim openssh-clients#支持重写rewrite,正则表达式
unzip pcre-8.33.zip
cd pcre-8.33
./configure --prefix=/usr/local/pcre
make && make install
cd ..#支持网页压缩
tar xf zlib-1.2.8.tar.gz
cd zlib-1.2.8
./configure  --prefix=/usr/local/zlib
make && make install
cd ..#支持https
tar xf openssl-1.0.0l.tar.gz
cd openssl-1.0.0l
./config --prefix=/usr/local/openssl
make && make install
cd ..#安装mp4和flv模块
tar xf nginx_mod_h264_streaming-2.2.7_update.tar.gz#创建运行nginx服务的账号和组
groupadd -r nginx
useradd -g nginx -r -s /sbin/nologin nginx#安装nginx
tar xf nginx-1.4.7.tar.gz
cd nginx-1.4.7
#启用flv和mp4功能
./configure \
--user=nginx \
--group=nginx \
--prefix=/usr/local/nginx \
--with-pcre=../pcre-8.33 \
--with-zlib=../zlib-1.2.8 \
--with-openssl=../openssl-1.0.0l \
--with-http_stub_status_module \
--pid-path=/var/run/nginx/nginx.pid  \
--lock-path=/var/lock/nginx.lock \
--with-http_gzip_static_module \
--http-proxy-temp-path=/var/tmp/nginx/proxy/ \
--http-fastcgi-temp-path=/var/tmp/nginx/fcgi/ \
--http-uwsgi-temp-path=/var/tmp/nginx/uwsgi \
--http-scgi-temp-path=/var/tmp/nginx/scgi \
--with-ipv6 \
--with-http_flv_module \
--with-http_mp4_module
make && make installcd ..#提供nginx启动脚本
cat > /etc/init.d/nginx << EOF
#!/bin/sh
#
# nginx - this script starts and stops the nginx daemin
#
# chkconfig:   - 85 15
# description:  Nginx is an HTTP(S) server, HTTP(S) reverse \
#               proxy and IMAP/POP3 proxy server
# processname: nginx
# config:      /usr/local/nginx/conf/nginx.conf
# pidfile:     /usr/local/nginx/logs/nginx.pid
# Source function library.
. /etc/rc.d/init.d/functions
# Source networking configuration.
. /etc/sysconfig/network
# Check that networking is up.
[ "\$NETWORKING" = "no" ] && exit 0
nginx="/usr/local/nginx/sbin/nginx"
prog=\$(basename \$nginx)
NGINX_CONF_FILE="/usr/local/nginx/conf/nginx.conf"
lockfile=/var/lock/subsys/nginx
start() { [ -x \$nginx ] || exit 5 [ -f \$NGINX_CONF_FILE ] || exit 6 echo -n \$"Starting \$prog: " daemon \$nginx -c \$NGINX_CONF_FILE
retval=\$? echo [ \$retval -eq 0 ] && touch \$lockfile return \$retval
}
stop() { echo -n \$"Stopping \$prog: " killproc \$prog -QUIT
retval=\$? echo [ \$retval -eq 0 ] && rm -f \$lockfile return \$retval
}
restart() { configtest || return \$? stop start
}
reload() { configtest || return \$? echo -n \$"Reloading \$prog: " killproc \$nginx -HUP
RETVAL=\$? echo
}
force_reload() { restart
}
configtest() { \$nginx -t -c \$NGINX_CONF_FILE
}
rh_status() { status \$prog
}
rh_status_q() { rh_status >/dev/null 2>&1
}
case "\$1" in start) rh_status_q && exit 0 \$1 ;; stop) rh_status_q || exit 0 \$1 ;; restart|configtest) \$1 ;; reload) rh_status_q || exit 7 \$1 ;; force-reload) force_reload ;; status) rh_status ;; condrestart|try-restart) rh_status_q || exit 0 ;; *) echo \$"Usage: \$0 {start|stop|status|restart|condrestart|try-restart|reload|force-reload|configtest}" exit 2
esac
EOF##赋予nginx执行的权限
chmod +x /etc/init.d/nginx
chkconfig --add nginx#创建文件
mkdir -p /var/tmp/nginx/proxy#启动nginx服务
/etc/init.d/nginx restart#查看nginx状态
/bin/ps -ef |grep nginx

2、分别在两台机器上创建不同的测试页面[为了测试]

#在master机器创建index.html页面
echo "Welcome To 192.168.0.189" > /usr/local/nginx/html/index.html
#在backup机器创建index.html页面
echo "Welcome To 192.168.0.200" > /usr/local/nginx/html/index.html

3、安装keepalived

#至于keepalived的安装可以选择源码编译安装或者yum源安装，都可以；这里选择源码编译安装

#安装keepalived源码安装的依赖包
yum -y install openssl openssl-devel
yum -y install popt-devel#源码编译安装keepalived
tar xf keepalived-1.2.7.tar.gz
cd keepalived-1.2.7
./configure --prefix=/usr/local/keepalived
make && make install#拷贝keepavlied的一些文件到相关目录中
mkdir /etc/keepalived
cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
cp /usr/local/keepalived/etc/rc.d/init.d/keepalived /etc/init.d/
cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/

修改keepalived配置文件

keepalived的文件路径/etc/keepalived/keepalived.conf

#####【Master-keepalived】#####

! Configuration File for keepalived   #全局定义  global_defs {notification_email {               #指定keepalived在发生事件时(比如切换),需要发送的email对象,可以多个,每行一个 zhengyansheng@cdvcloud.com}notification_email_from admin@cdvcloud.comsmtp_server 127.0.0.1              #指定发送email的smtp服务器smtp_connect_timeout 30router_id LVS_DEVEL                #运行keepalived的机器的一个标识
}vrrp_instance VI_1 {state MASTER               #为主服务器interface eth1             #监听的本地网卡接口virtual_router_id 51       #主辅virtual_router_id号必须相同mcast_src_ip=192.168.0.189 #主nginx的ip地址priority 100               #优先级为100，此值越大优先级越大 就为master 权重值advert_int 1               #VRRP Multicast 广播周期秒数；心跳检测时间，单位秒authentication {auth_type PASS         #vrrp认证方式auth_pass 1111         #vrrp口令}virtual_ipaddress {        #VRRP HA 虚拟地址 如果有多个VIP，继续换行填写192.168.0.250/24 dev eth1}
}

#####【Backup-keepalived】#####

! Configuration File for keepalivedglobal_defs {notification_email {zhengyansheng@cdvcloud.com}notification_email_from admin@cdvcloud.comsmtp_server 127.0.0.1smtp_connect_timeout 30router_id LVS_DEVEL
}vrrp_instance VI_1 {state BACKUPinterface eth1virtual_router_id 51mcast_src_ip=192.168.0.200priority 99advert_int 1authentication {auth_type PASSauth_pass 1111}virtual_ipaddress {192.168.0.250/24 dev eth1}
}

启动keepalived服务[所有的服务开机启动]

/etc/init.d/keepalived start

chkconfig nginx on

chkconfig keepalived on

查看vip状态

#首先在master节点上查看vip的状态

[root@master ~]# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00inet 127.0.0.1/8 scope host loinet6 ::1/128 scope host valid_lft forever preferred_lft forever
2: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000link/ether 00:0c:29:59:45:de brd ff:ff:ff:ff:ff:ffinet 192.168.0.189/24 brd 192.168.0.255 scope global eth1inet 192.168.0.250/32 scope global eth1          #vip地址已经加入到master节点上inet6 fe80::20c:29ff:fe59:45de/64 scope link valid_lft forever preferred_lft forever

#其次在backup节点上查看vip的状态

[root@backup ~]# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00inet 127.0.0.1/8 scope host loinet6 ::1/128 scope host valid_lft forever preferred_lft forever
2: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000link/ether 00:0c:29:61:70:d3 brd ff:ff:ff:ff:ff:ffinet 192.168.0.200/24 brd 192.168.0.255 scope global eth1inet6 fe80::20c:29ff:fe61:70d3/64 scope link valid_lft forever preferred_lft forever

学会看日志输出

一部分：当启动keepalived服务的时候，会根据配置文件的优先级来竞选谁为master，从日志来看192.168.0.189竞选master

[root@master ~]# /etc/init.d/keepalived start
正在启动 keepalived：                                      [确定][root@master ~]# tail -f /var/log/messages
Aug 16 11:09:48 localhost Keepalived[25707]: Starting Keepalived v1.2.7 (08/16,2014)
Aug 16 11:09:48 localhost Keepalived[25708]: Starting Healthcheck child process, pid=25710
Aug 16 11:09:48 localhost Keepalived[25708]: Starting VRRP child process, pid=25711
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Interface queue is empty
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Netlink reflector reports IP 192.168.0.189 added
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Netlink reflector reports IP fe80::20c:29ff:fe59:45de added
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Registering Kernel netlink reflector
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Registering Kernel netlink command channel
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Registering gratuitous ARP shared channel
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Interface queue is empty
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Netlink reflector reports IP 192.168.0.189 added
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Netlink reflector reports IP fe80::20c:29ff:fe59:45de added
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Registering Kernel netlink reflector
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Registering Kernel netlink command channel
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Opening file '/etc/keepalived/keepalived.conf'.
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Configuration is using : 62766 Bytes
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: Using LinkWatch kernel netlink reflector...
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Opening file '/etc/keepalived/keepalived.conf'.
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Configuration is using : 7073 Bytes
Aug 16 11:09:48 localhost Keepalived_vrrp[25711]: VRRP sockpool: [ifindex(2), proto(112), fd(11,12)]
Aug 16 11:09:48 localhost Keepalived_healthcheckers[25710]: Using LinkWatch kernel netlink reflector...
Aug 16 11:09:51 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) Transition to MASTER STATE
Aug 16 11:09:56 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) Entering MASTER STATE
Aug 16 11:09:56 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) setting protocol VIPs.
Aug 16 11:09:56 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth1 for 192.168.0.250
Aug 16 11:09:56 localhost Keepalived_healthcheckers[25710]: Netlink reflector reports IP 192.168.0.250 added
Aug 16 11:10:01 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth1 for 192.168.0.250[root@backup ~]# tail -f /var/log/messages
Aug 16 11:09:51 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Received higher prio advert
Aug 16 11:09:51 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Entering BACKUP STATE
Aug 16 11:09:51 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) removing protocol VIPs.
Aug 16 11:09:51 localhost Keepalived_healthcheckers[26135]: Netlink reflector reports IP 192.168.0.250 removed[root@master ~]# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00inet 127.0.0.1/8 scope host loinet6 ::1/128 scope host valid_lft forever preferred_lft forever
2: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000link/ether 00:0c:29:59:45:de brd ff:ff:ff:ff:ff:ffinet 192.168.0.189/24 brd 192.168.0.255 scope global eth1inet 192.168.0.250/32 scope global eth1inet6 fe80::20c:29ff:fe59:45de/64 scope link valid_lft forever preferred_lft forever二部分：当我们停掉主keepalived服务的时候，主master的vip会自动切换到辅机器上，看日志
停止主keepalived服务
[root@master ~]# /etc/init.d/keepalived stop
停止 keepalived：                                          [确定][root@master ~]# tail -f /var/log/messages
Aug 16 11:16:14 localhost Keepalived[25708]: Stopping Keepalived v1.2.7 (08/16,2014)
Aug 16 11:16:14 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) sending 0 priority
Aug 16 11:16:14 localhost Keepalived_vrrp[25711]: VRRP_Instance(VI_1) removing protocol VIPs.[root@backup ~]# tail -f /var/log/messages
Aug 16 11:09:51 localhost Keepalived_healthcheckers[26135]: Netlink reflector reports IP 192.168.0.250 removed
Aug 16 11:16:14 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Transition to MASTER STATE
Aug 16 11:16:19 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Entering MASTER STATE
Aug 16 11:16:19 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) setting protocol VIPs.
Aug 16 11:16:19 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth1 for 192.168.0.250
Aug 16 11:16:19 localhost Keepalived_healthcheckers[26135]: Netlink reflector reports IP 192.168.0.250 added
Aug 16 11:16:24 localhost Keepalived_vrrp[26136]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth1 for 192.168.0.250[root@backup ~]# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00inet 127.0.0.1/8 scope host loinet6 ::1/128 scope host valid_lft forever preferred_lft forever
2: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000link/ether 00:0c:29:61:70:d3 brd ff:ff:ff:ff:ff:ffinet 192.168.0.200/24 brd 192.168.0.255 scope global eth1inet 192.168.0.250/32 scope global eth1  辅节点已经成功接管vipinet6 fe80::20c:29ff:fe61:70d3/64 scope link valid_lft forever preferred_lft forever

测试1：nginx主备切换的时间
C:\Users\Allentuns>ping 192.168.0.250 -t正在 Ping 192.168.0.250 具有 32 字节的数据:
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
请求超时。
请求超时。
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64
来自 192.168.0.250 的回复: 字节=32 时间<1ms TTL=64192.168.0.250 的 Ping 统计信息:数据包: 已发送 = 28，已接收 = 26，丢失 = 2 (7% 丢失)，
往返行程的估计时间(以毫秒为单位):最短 = 0ms，最长 = 0ms，平均 = 0ms
Control-C
^C
C:\Users\Allentuns>

以上keepalived的配置文件并不能满足整个架构：

因为默认情况下，keepalived工作模式并不能直接监控nginx服务，只有当keepalived服务挂掉后才能主备切换，nginx服务挂掉后不能实现主备服务器的切换，但是我们的目的就是要实现nginx服务keepalived挂掉后，都要主备切换。

以上有两种方法可以实现

1. keepalived配置文件中可以支持shell脚本，写个监听nginx服务的脚本就可以了

2. 单独写个脚本来监听nginx和keepalived服务

基于第一种情况

#####【Master-keepalived】#####

! Configuration File for keepalived   #全局定义  global_defs {notification_email {               #指定keepalived在发生事件时(比如切换),需要发送的email对象,可以多个,每行一个 zhengyansheng@cdvcloud.com}notification_email_from admin@cdvcloud.comsmtp_server 127.0.0.1              #指定发送email的smtp服务器smtp_connect_timeout 30router_id LVS_DEVEL                #运行keepalived的机器的一个标识
}vrrp_script
chk_nginx {               #检测nginx服务是否在运行有很多方式，比如进程，用脚本检测等等 script "killall -0 nginx"  #用shell命令检查nginx服务是否存在 interval 1                 #时间间隔为1秒检测一次 weight -2                  #当nginx的服务不存在了，就把当前的权重-2 fall 2                     #测试失败的次数 rise 1                     #测试成功的次数
} vrrp_instance VI_1 {state MASTER               #为主服务器interface eth1             #监听的本地网卡接口virtual_router_id 51       #主辅virtual_router_id号必须相同mcast_src_ip=192.168.0.189 #主nginx的ip地址priority 100               #优先级为100，此值越大优先级越大 就为master 权重值advert_int 1               #VRRP Multicast 广播周期秒数；心跳检测时间，单位秒authentication {auth_type PASS         #vrrp认证方式auth_pass 1111         #vrrp口令}virtual_ipaddress {        #VRRP HA 虚拟地址 如果有多个VIP，继续换行填写192.168.0.250/24 dev eth1}track_script { chk_nginx   #引用上面的vrrp_script定义的脚本名称 }
}

#####【Backup-keepalived】#####

! Configuration File for keepalivedglobal_defs {notification_email {zhengyansheng@cdvcloud.com}notification_email_from admin@cdvcloud.comsmtp_server 127.0.0.1smtp_connect_timeout 30router_id LVS_DEVEL
}vrrp_script
chk_nginx {               #检测nginx服务是否在运行有很多方式，比如进程，用脚本检测等等 script "killall -0 nginx"  #用shell命令检查nginx服务是否存在 interval 1                 #时间间隔为1秒检测一次 weight -2                  #当nginx的服务不存在了，就把当前的权重-2 fall 2                     #测试失败的次数 rise 1                     #测试成功的次数
} vrrp_instance VI_1 {state BACKUPinterface eth1virtual_router_id 51mcast_src_ip=192.168.0.200priority 99advert_int 1authentication {auth_type PASSauth_pass 1111}virtual_ipaddress {192.168.0.250/24 dev eth1}track_script { chk_nginx   #引用上面的vrrp_script定义的脚本名称 }
}#这种情况的监听脚本是2秒中监听一次nginx服务，如果进程不存在，则将keepalived的权重减少2，为98，那么权重为99的为主 自动切换到备用服务器上

基于第二种情况

#!/bin/bash
while  :
do nginxpid=`ps -C nginx --no-header | wc -l` if [ $nginxpid -eq 0 ];then /usr/local/nginx/sbin/nginx sleep 5 nginxpid=`ps -C nginx --no-header | wc -l` echo $nginxpid if [ $nginxpid -eq 0 ];then /etc/init.d/keepalived stop fi fi sleep 5
done#这个脚本确实很好用，来自酒哥的博文

至于后端tomcat的负载均衡和主备切换在这里就不多说了！很简单。。。。 网上有很多这方面的教程

我会在下周末之前把nginx的双主架构和hadoop的源码编译贡献给博客上。

博文的笔记和所有的软件都已经上传！期待和大家交流更多linux运维 QQ 467754239

http://yunpan.cn/QaSn4mzCd65YP  访问密码 06b3

----------补充时间 2015-01-14----------

主Nginx机器之一的Keepalived.conf配置文件如下：

! Configuration File for keepalived
global_defs {notification_email {yuhongchun027@163.com}notification_email_from keepalived@chtopnet.comsmtp_server 127.0.0.1smtp_connect_timeout 30router_id LVS_DEVEL
}
vrrp_instance VI_1 {state MASTERinterface eth0virtual_router_id 51priority 100advert_int 1authentication {auth_type PASSauth_pass 1paituan.com}virtual_ipaddress {192.168.1.8}
}
vrrp_instance VI_2 {state BACKUPinterface eth0virtual_router_id 52priority 99advert_int 1authentication {auth_type PASSauth_pass 1paituan.com}virtual_ipaddress {192.168.1.9}
}

主Nginx之二的keepalivd.conf配置文件如下：

! Configuration File for keepalived
global_defs {notification_email {yuhongchun027@163.com}notification_email_from keepalived@chtopnet.comsmtp_server 127.0.0.1smtp_connect_timeout 30router_id LVS_DEVEL
}
vrrp_instance VI_1 {state BACKUPinterface eth0virtual_router_id 51priority 99advert_int 1authentication {auth_type PASSauth_pass 1paituan}virtual_ipaddress {192.168.1.8                 }
}
vrrp_instance VI_2 {state MASTERinterface eth0virtual_router_id 52priority 100advert_int 1authentication {auth_type PASSauth_pass 1paituan}virtual_ipaddress {192.168.1.9                 }
}

二台机器的监控Nginx的进程脚本，脚本内容如下：

#!/bin/bash
while  :
do
nginxpid=`ps -C nginx --no-header | wc -l`if [ $nginxpid -eq 0 ];then/usr/local/nginx/sbin/nginxsleep 5
nginxpid=`ps -C nginx --no-header | wc -l`echo $nginxpidif [ $nginxpid -eq 0 ];then/etc/init.d/keepalived stopfifisleep 5
done

nohup sh /root/nginxpid.sh &

此脚本我是直接从生产服务器上下载的，大家不要怀疑它会引起死循环和有效性的问题，我稍为解释一下，这是一个无限循环的脚本，放在主Nginx机器上（因为目前主要是由它提供服务），每隔5秒执行一次，用ps -C 命令来收集nginx的PID值到底是否为0，如果是0的话（即Nginx进程死掉了），尝试启动nginx进程；如果继续为0，即nginx启动失改， 则关闭本机的Keeplaived进程，VIP地址则会由备机接管，当然了，整个网站就会由备机的Nginx来提供服务了，这样保证Nginx进程的高可用。

最近工作不是很忙，花时间做了个实验，来验证vip能不能是公网地址中的一个或者多个。

验证的结果是这样的：如果按照上图来分配地址，使用公网地址117.25.36.41和117.25.36.43中的任意一个来充当vip的后果是，物理网卡上eth1上的地址就会失效，导致无法使用xshell的ssh连接服务器，因为找不到地址了。只有vip浮动到那个服务器上才能远程登录到服务器上。

建议：vip最好不使用服务器上绑定的物理网卡，以免影响不必要的麻烦。

时间：2015-04-13 补充

Apr 13 17:08:07 localhost Keepalived_vrrp[2186]: bogus VRRP packet received on eth0 !!!
Apr 13 17:08:07 localhost Keepalived_vrrp[2186]: VRRP_Instance(VI_1) ignoring received advertisment...
Apr 13 17:08:08 localhost Keepalived_vrrp[2186]: ip address associated with VRID not present in received packet : 192.168.0.251
Apr 13 17:08:08 localhost Keepalived_vrrp[2186]: one or more VIP associated with VRID mismatch actual MASTER advert
Apr 13 17:08:08 localhost Keepalived_vrrp[2186]: bogus VRRP packet received on eth0 !!!
Apr 13 17:08:08 localhost Keepalived_vrrp[2186]: VRRP_Instance(VI_1) ignoring received advertisment...

解决办法：

改变配置文件/etc/keepalived/keepalived.conf 中virtual_route_id的值

比如virtual_router_id 60 主从方都要改，默认为51