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。
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。二者结合，可以构架出比较稳定的软件lb方案。
现在有两台虚拟机ServerA和ServerB. 两个对外提供Web服务器的虚IP(VIP)192.168.200.100和192.168.200.200, 虚IP用在keepalived的配置中, 网卡接口配置有内网IP.
ServerA: eth0: 192.168.200.128 VIP: 192.168.200.100 （www.srt.com.cn） ServerB: eth0: 192.168.200.129 VIP: 192.168.200.200 （www.srtedu.com）
如果两台服务器都正常地提供网络服务, 那么, 发往192.168.200.100的服务请求会被ServerA处理, 发往192.168.200.200的服务请求会被ServerB处理. 假设只有ServerB出现故障, 那么, 所有的请求都由ServerA进行处理. 当只有ServerA出现故障时, 也是同理.
软件环境：本文使用的Linux为RHEL 5.2、软件：keepalived-1.1.19、nginx-0.7.64
安装Keepalived
wget http://www.keepalived.org/software/keepalived-1.1.19.tar.gz
tar zxvf keepalived-1.1.19.tar.gz
cd keepalived-1.1.19
./configure --prefix=/usr/local/keepalived
make
make install
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/
mkdir /etc/keepalived
cd /etc/keepalived/
配置Keepalived
接下来, 是最重要的修改配置文件/etc/keepalived/keepalived.conf.
global_defs { notification_email { [email protected] } notification_email_from [email protected] smtp_server 211.155.225.210 smtp_connect_timeout 30 router_id srtweb
}
vrrp_instance VI_1 { state MASTER interface eth0 virtual_router_id 51 priority 100 advert_int 1 smtp_alert authentication { auth_type PASS auth_pass srt_L7switch } virtual_ipaddress { 192.168.200.100 }}
vrrp_instance VI_2 { state BACKUP interface eth0 virtual_router_id 52 priority 50 advert_int 1 smtp_alert authentication { auth_type PASS auth_pass srt_L7switch } virtual_ipaddress { 192.168.200.200 }}
执行 /etc/rc.d/init.d/keepalived start 启动keepalived后, 执行ip a, 你将看到类似的信息:
[[email protected] ~]# ip a 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:01:11:2a brd ff:ff:ff:ff:ff:ff inet 192.168.200.128/24 brd 192.168.200.255 scope global eth0 inet 192.168.200.100/32 scope global eth0 inet 192.168.200.200/32 scope global eth0 inet6 fe80::20c:29ff:fe01:112a/64 scope link
可以看到, ServerA的网卡绑定了两个虚IP 192.168.200.100和192.168.200.200, 这时在第3台机器上ping这两个IP, 是可以通的.
然后, 按上面的方法安装ServerB. ServerB的keepalived.conf配置和ServerA基本相同, 但是, 把state MASTER和state BACKUP调换, priority 100和priority 50调换,router_id改为srtedu. 启动ServerB后, 再到ServerA上运行ip a, 你将看到
[[email protected] ~]# ip a 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:01:11:2a brd ff:ff:ff:ff:ff:ff inet 192.168.200.128/24 brd 192.168.200.255 scope global eth0 inet 192.168.200.100/32 scope global eth0 inet6 fe80::20c:29ff:fe01:112a/64 scope link
192.168.200.200已经不见了, 因为被ServerB使用了, ServerB是这个IP的MASTER, 它有优先使用权. 这样, 两个IP的网络数据分别被两台服务器处理. 现在验证ServerB在出故障的情况, 将ServerB的网线拔掉, 然后在ServerA上执行ip a, 你将看到
[[email protected] ~]# ip a 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:01:11:2a brd ff:ff:ff:ff:ff:ff inet 192.168.200.128/24 brd 192.168.200.255 scope global eth0 inet 192.168.200.100/32 scope global eth0 inet 192.168.200.200/32 scope global eth0 inet6 fe80::20c:29ff:fe01:112a/64 scope link
192.168.200.200又被ServerA使用了.
安装Nginx
1、创建供Nginx使用的组和帐号：
/usr/sbin/groupadd www -g 48
/usr/sbin/useradd -u 48 -g www www
2、编译安装rewrite模块支持包
wget ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre/pcre-7.7.tar.gz
tar zxvf pcre-7.7.tar.gz
cd pcre-7.7/
./configure
make && make install
cd ../
3、编译安装Nginx
wget http://sysoev.ru/nginx/nginx-0.7.64.tar.gz
tar zxvf nginx-0.7.64.tar.gz
cd nginx-0.7.64/
./configure --user=www --group=www --prefix=/usr/local/nginx --with-http_stub_status_module --with-http_ssl_module --with-http_realip_module --with-http_flv_module
make && make install
cd ../
4、备份默认nginx.conf配置文件
mv /usr/local/nginx/conf/nginx.conf /usr/local/nginx/conf/nginx.old
5、创建Nginx配置文件
#vi /usr/local/nginx/conf/nginx.conf
user www www;
worker_processes 8;
pid /var/run/nginx.pid;
worker_rlimit_nofile 51200;
events
{
use epoll;
worker_connections 51200;
}
http
{
include mime.types;
default_type application/octet-stream;
charset gb2312;
server_names_hash_bucket_size 128;
client_header_buffer_size 32k;
large_client_header_buffers 4 32k;
sendfile on;
tcp_nopush on;
keepalive_timeout 60;
tcp_nodelay on;
gzip on;
gzip_min_length 1k;
gzip_buffers 4 16k;
gzip_http_version 1.0;
gzip_comp_level 2;
gzip_types text/plain application/x-javascript text/css application/xml;
gzip_vary on;
upstream srtweb {
server 192.168.200.201:80;
server 192.168.200.202:80;
server 192.168.200.203:80;
}
upstream srtedu {
server 192.168.200.211:80;
server 192.168.200.212:80;
server 192.168.200.213:80;
}
server {
listen 80;
server_name www.srt.com.cn srt.com.cn;
location /{
proxy_pass http://srtweb;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
access_log /var/log/access_srtweb.log combined;
}
server {
listen 80;
server_name www.srtedu.com srtedu.com;
location /{
proxy_pass http://srtedu;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
access_log /var/log/access_srtedu.log combined;
}
}
#第1个server同时也作为默认主机
6、启动
ulimit -SHn 102400
/usr/local/nginx/sbin/nginx
7、同样的在Server B上进行1~6步骤的安装配置Nginx操作。
后记： 对于中、小型企业，如果没有资金去购买昂贵的四/七层负载均衡交换机，那么Nginx是不错的七层负载均衡选择，并且可以通过Nginx + Keepalived实现Nginx 负载均衡器双机互备，任意一台机器发生故障，对方都能够将虚拟IP接管过去。
这个环境用webbench做压力测试，在10000并发的情况下
下载地址：http://blog.s135.com/soft/linux/webbench/webbench-1.5.tar.gz
[[email protected] webbench-1.5]# webbench -c 10000 -t 10 http://www.srt.com.cn/
Webbench - Simple Web Benchmark 1.5
Copyright (c) Radim Kolar 1997-2004, GPL Open Source Software.
Benchmarking: GET http://www.srt.com.cn/
10000 clients, running 10 sec.
Speed=746202 pages/min, 2785675 bytes/sec.
Requests: 124331 susceed, 36 failed
由于其中之一的nginx主机只有64M内存，所以不敢用高于10000的并发测试，如果同志们想用于生产环境，可考虑更高的并发测试(webbench最多支持30000并发)