Nginx_Master: 192.168.1.103 提供负载均衡
Nginx_BackUp: 192.168.1.104 负载均衡备机
Nginx_VIP_TP: 192.168.1.108 网站的 VIP 地址(虚拟 IP)
原理：
VIP 是外网访问的IP地址，通过 keepalived 设置，以及 VRRP 将 VIP 绑定到主机和备机上，通过权重实现控制。当主机挂掉后，keepalived 释放对主机的控制，备机接管VIP。
扩展：
主机和备机可进一步延伸，它们只通过 Nginx 提供负载均衡。再让其它的机器提供真实的 web 服务。
**********************************************************************************************
安装 Nginx (省略)
-----------------------
Keepalived 安装
-----------------------
wget http://www.keepalived.org/software/keepalived-1.2.7.tar.gz
tar -zxf keepalived-1.2.7.tar.gz
cd keepalived-1.2.7
64 位系统：
./configure --sysconf=/etc --prefix=/usr/local/keepalived --with-kernel-dir=/usr/src/kernels/2.6.32-358.2.1.el6.x86_64/
或 32 位系统:
./configure --sysconf=/etc --prefix=/usr/local/keepalived --with-kernel-dir=/usr/src/kernels/2.6.32-358.6.2.el6.i686/
可通过 getconf LONG_BIT 得到系统位数。
参数解释:
--sysconf 指定了配置文件的地址.即：/etc/keepalived/keepalived.conf
--prefix 指定了安装目录
--with-kernel-dir 指定使用内核源码中的头文件,即 include 目录.只有使用 LVS 时才需要这个参数，其它的时候不需要。
如果报错:
configure: error: Popt libraries is required
解决:
yum install popt-devel
再 configue .成功后提示：
Keepalived configuration
------------------------
Keepalived version : 1.2.7
Compiler : gcc
Compiler flags : -g -O2
Extra Lib : -lpopt -lssl -lcrypto
Use IPVS Framework : Yes
IPVS sync daemon support : Yes
IPVS use libnl : No
Use VRRP Framework : Yes
Use VRRP VMAC : Yes
SNMP support : No
Use Debug flags : No
安装：
make
make install
设置成为服务并开机启动:
cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
/etc/rc.d/init.d/keepalived status
chkconfig --add keepalived
chkconfig keepalived on
设置主机上的配置文件内容：
vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
}
notification_email_from [email protected]
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id LVS_DEVEL
}
vrrp_instance VI_1 {
state MASTER
interface eth0
virtual_router_id 51
# 此处是主 Nginx 的 IP 地址.
mcast_src_ip
192.168.1.103
# 该机的 priority(优先) 为 100
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 1111222
}
virtual_ipaddress {
192.168.1.108
}
}
前面的结构那里已经规定好了 VIP 和 主备机的 IP, 所以这里按上面的填。
备机的配置文件:
! Configuration File for keepalived
global_defs {
notification_email {
}
notification_email_from [email protected]
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id LVS_DEVEL
}
vrrp_instance VI_1 {
state SLAVER
interface eth0
virtual_router_id 51
# 此处是备 Nginx 的 IP 地址.
mcast_src_ip 192.168.1.104
# 该机的 priority(优先) 为 99
priority 99
advert_int 1
authentication {
auth_type PASS
auth_pass 1111222
}
virtual_ipaddress {
192.168.1.108
}
}
这时候 ping 192.168.1.108 是不通的.
然后在两台机器上分别启动 keepalived 服务
这时候再 ping 192.168.1.108 .通了.
实际上这时候 108 是被绑到主机上的。在主机上：
查看系统日志:
tailf /var/log/messages
May 29 18:32:16 localhost Keepalived_vrrp[27731]: Opening file '/etc/keepalived/keepalived.conf'.
May 29 18:32:16 localhost Keepalived_vrrp[27731]: Configuration is using : 62906 Bytes
May 29 18:32:16 localhost Keepalived_vrrp[27731]: Using LinkWatch kernel netlink reflector...
May 29 18:32:16 localhost Keepalived_healthcheckers[27729]: Using LinkWatch kernel netlink reflector...
May 29 18:32:16 localhost Keepalived_vrrp[27731]: VRRP sockpool: [ifindex(2), proto(112), fd(11,12)]
May 29 18:32:17 localhost Keepalived_vrrp[27731]: VRRP_Instance(VI_1) Transition to MASTER STATE
May 29 18:32:18 localhost Keepalived_vrrp[27731]: VRRP_Instance(VI_1) Entering MASTER STATE
May 29 18:32:18 localhost Keepalived_vrrp[27731]: VRRP_Instance(VI_1) setting protocol VIPs.
May 29 18:32:18 localhost Keepalived_vrrp[27731]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.1.108
May 29 18:32:18 localhost Keepalived_healthcheckers[27729]: Netlink reflector reports IP 192.168.1.108 added
可以看到.
VRRP(虚拟路由冗余协议)已经启动.我们可以通过命令 ip addr 来检查主 Nginx 上的 IP 分配情况.
[root@localhost ~]# ip addr
1: lo: 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: mtu 1500 qdisc mq state UP qlen 1000
link/ether 00:15:c5:ef:53:8c brd ff:ff:ff:ff:ff:ff
inet 192.168.1.103/25 brd 192.168.1.255 scope global eth0
inet 192.168.1.108/32 scope global eth0
inet6 fe80::215:c5ff:feef:538c/64 scope link
valid_lft forever preferred_lft forever
3: eth1: mtu 1500 qdisc noop state DOWN qlen 1000
link/ether 00:15:c5:ef:53:8e brd ff:ff:ff:ff:ff:ff
可以看到 VIP 地址已经绑定到主 Nginx 机器上: inet 192.168.1.108/32 scope global eth0
我们通过 tcpdump 抓包：
[root@localhost ~]# tcpdump vrrp
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes
13:38:27.797982 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:28.794693 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:29.794518 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:30.798581 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:31.795902 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:32.804050 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:33.801191 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
13:38:34.798793 IP htuidc.bgp.ip > vrrp.mcast.net: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
这样，一个 Nginx + Keepalived 的架构就完成了。
监控和主备切换
接下来可以完善一下，加上实时监控，如果发现负载均衡的 Nginx 出现问题，就将该机器上的 Keepalived 服务停掉。
nginx_check.sh:
#!/bin/bash
while :
do
nginxpid = 'ps -C nginx --no-header | wc -l'
if[ $nginxpid -eq 0 ];then
service nginx start
sleep 3
nginxpid = 'ps -C nginx --no-header | wc -l'
echo $nginxpid
if[ $nginxpid -eq 0 ];then
service keepalived stop
fi
fi
sleep 3
done
然后让该脚本一直在后台运行:
nohup /etc/nginx_check.sh
或者将它添加成服务，让它开机自启动:
测试:
--------------------
在两台机器的 web 服务器上分别放一个 index.html, 里面内容分别是自己机器的IP.
通过VIP访问：
http://192.168.1.108/index.html 发现显示的是主机的IP.
此时，关掉主机的 nginx, 这时候由于上面的监控脚本。主机的 keepalived 也会关闭。这时候再访问上面地址，发现显示的是备机的IP。可见，切换成功。
负载均衡
--------------------
通过 ip_hash 实现会话保持
原文地址:http://blog.sina.com.cn/s/blog_5f54f0be0101eyff.html