利用nginx的url_hash提高squid的缓存效率

nginx的url_hash模块（第三方）
含义：按访问url的hash结果来分配请求，使每个url定向到同一个后端服务器，后端服务器为缓存时比较有效。
nginx本身是不支持url_hash的，需要打第三方的patch，下载地址：
http://anotherbug.blog.chinajavaworld.com/servlet/AttachmentServlet/download/123297-4298-2148/nginx_upstream_hash-0[1].2.tar.gz
See http://wiki.codemongers.com/NginxHttpUpstreamRequestHashModule for more details.
关于nginx的url_hash模块的解释就到此为止，下面我们来看看怎么用nginx的url_hash模块来提高squid的缓存效率的：

一.实验环境
四台服务器，主机名和IP分别为：
192.168.4.191----------nginx
192.168.4.188----------squid01
192.168.4.190----------squid02
192.168.4.189----------真实web

二.nginx和squid的安装
1.nginx以及nginx的url_hash模块的安装
wget http://anotherbug.blog.chinajavaworld.com/servlet/AttachmentServlet/download/123297-4298-2148/nginx_upstream_hash-0[1].2.tar.gz
wget http://www.nginx.eu/download/sources/nginx-0.5.21.tar.gz
1) tar -zxvf nginx_upstream_hash-0.2.tar.gz -C /usr/local
   mv nginx_upstream_hash-0.2 nginx
2) tar -zxvf nginx-0.5.21.tar.gz
   cd  nginx-0.5.21
   patch -p0 < /usr/local/nginx/nginx-0.5.21.patch
   ./configure --with-http_stub_status_module --add-module=/usr/local/nginx/
   make && make install
3) cd /usr/local/nginx/conf/
   vi nginx.conf 内容如下：

user  nobody nobody;
worker_processes 4;
#error_log  logs/error.log;
#error_log  logs/error.log  notice;
#error_log  logs/error.log  info;
#pid        logs/nginx.pid;
worker_rlimit_nofile 51200;
events
{
      use epoll;
      worker_connections 51200;
}

http
{
#      include       mime.types;
      default_type  application/octet-stream;
      keepalive_timeout 20;
      tcp_nodelay on;
      upstream  www.jialisong.com  {
                server   192.168.4.188 ;
                server   192.168.4.190 ;
                hash $request_uri;
      }
server
      {
              listen      80;
              server_name www.jialisong.com;
              location /         {
                       proxy_pass        http://www.jialisong.com;
                       proxy_set_header   Host             $host;
                       proxy_set_header   X-Forwarded-For  $proxy_add_x_forwarded_for;
                       proxy_set_header   X-Real-IP        $remote_addr;
        }
              log_format  main  '$remote_addr - $remote_user [$time_local] $request '
                                '"$status" $body_bytes_sent "$http_referer" '
                                '"$http_user_agent" "$http_x_forwarded_for"';
              access_log  /usr/local/nginx/logs/access_log  ;
             location  /NginxStatus  {
                stub_status            on;
                access_log              on;
                auth_basic              "NginxStatus";
#               auth_basic_user_file  htpasswd;

                }
      }
 }

4） /usr/local/nginx/sbin/nginx -t  测试nginx的配置文件
    /usr/local/nginx/sbin/nginx     启动nginx

2.squid服务器的安装
1）tar -jxvf squid-2.7.STABLE5.tar.bz2
   cd squid-2.7.STABLE5
   ./configure  --prefix=/usr/local/squid --enable-dlmalloc --with-pthreads --enable-poll --disable-internal-dns --enable-stacktrace --enable-removal-policies="heap,lru" --enable-delay-pools --enable-storeio="aufs,coss,diskd,ufs"
   make && make install
   useradd -s /sbin/login squid
   chown -R  squid:squid  /usr/local/squid/var/
   mkdir /tmp1
   chmod -R 777 /tmp1
2) cd /usr/local/squid/etc/
   vi squid.conf  
内容如下：
acl all src 0.0.0.0/0.0.0.0
acl manager proto cache_object
acl Localhost src 127.0.0.1 192.168.4.188 192.168.4.191
acl to_localhost dst 127.0.0.0/8
acl SSL_ports port 443
acl Safe_ports port 80
acl CONNECT method CONNECT
#acl SNMP snmp_community   public
#acl myip dst 192.168.4.191
#http_access deny !myip
http_access allow all
http_access allow manager localhost
http_access deny manager
http_access deny !Safe_ports
http_access deny CONNECT !SSL_ports
#snmp_access allow SNMP Localhost
#acl myip dst 192.168.30.93
#http_access deny !myip

#acl Manager proto cache_object
#acl Localhost src 127.0.0.1 192.168.30.93
#http_access allow Manager Localhost
#http_access deny Manager

#acl Safe_ports port  81
#http_access deny !Safe_ports
#http_access allow all

#acl OverConnLimit maxconn 16
#http_access deny OverConnLimit
http_access deny all

icp_access allow all

http_port 192.168.4.188:80 vhost vport

cache_peer 192.168.4.189 parent 80 0 no-digest  no-query originserver name=www
cache_peer_domain www www.jialisong.com

hierarchy_stoplist cgi-bin ?
acl QUERY urlpath_regex cgi-bin .php .cgi .avi .wmv .rm .ram .mpg .mpeg .zip .exe
cache deny QUERY
cache_mem 300 MB
maximum_object_size_in_memory 4096 KB
cache_dir ufs /tmp1 2048 16 512
ie_refresh on
cache_swap_low 90
cache_swap_high 95
#ipcache_size 4096
#ipcache_low 90
#ipcache_high 95
fqdncache_size 1024
logformat combined %>a %ui %un [%tl] "%rm %ru HTTP/%rv" %Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
#LogFormat "%{X-Forwarded-For}i %l %u %t \"%r\" %>s %b \"%{Referer}i\" \"%{User-Agent}i\"" combined
access_log none
access_log /var/log/squid/access.log squid
cache_log /var/log/squid/cache.log squid
cache_store_log none
emulate_httpd_log on
request_header_max_size 100 KB
url_rewrite_host_header off

refresh_pattern ^ftp:           1440    20%     10080
refresh_pattern ^gopher:        1440    0%      1440
refresh_pattern .               0       20%     4320

refresh_pattern -i .gif$   1440    50%     2880      ignore-reload
refresh_pattern -i .jpg$   1440    50%     2880      ignore-reload
refresh_pattern -i .png$   1440    50%     2880      ignore-reload

refresh_pattern -i .mp3$   1440    50%     2880      ignore-reload
refresh_pattern -i .wmv$   1440    50%     2880      ignore-reload
refresh_pattern -i .rm$   1440    50%     2880      ignore-reload
refresh_pattern -i .swf$   1440    50%     2880      ignore-reload
refresh_pattern -i .mpeg$   1440    50%     2880      ignore-reload
refresh_pattern -i .wma$   1440    50%     2880      ignore-reload
refresh_pattern -i .bmp$  1440     50%     2880      ignore-reload
refresh_pattern -i .txt$  1440    50%     2880      ignore-reload

refresh_pattern -i .css$ 10 50% 60 reload-into-ims
refresh_pattern -i .js$ 10 50% 60 reload-into-ims
refresh_pattern -i .xml$ 10 50% 30 reload-into-imsfresh_pattern -i .html$ 720 50% 1440 reload-into-ims

acl apache rep_header Server ^Apache
broken_vary_encoding allow apache
cache_effective_user squid
cache_effective_group squid
visible_hostname pic.mytophome.com
cache_mgr [email protected]
httpd_suppress_version_string on
cache_vary on
icp_port 0
error_directory /usr/local/squid/share/errors/English
via off
forwarded_for on
tcp_recv_bufsize 65535 bytes
coredump_dir /usr/local/squid/var/cache
max_filedesc 20480

3)启动squid
/usr/local/squid/sbin/squid -z
/usr/local/squid/sbin/squid -NCd1
/usr/local/squid/sbin/squid -f /usr/local/squid/etc/squid.conf
两台squid的安装配置一样只是在squid.conf配置文件做稍微的修改即可

3.真实web的应用
服务器192.18.4.189为web服务器，通过访问:
http://192.168.4.189
能正常的访问即OK！！！ 

三.测试
1）修改自己电脑的hosts文件，把 www.jialisong.com指向到nginx服务器的ip上（目的在于使 www.jialisong.com和192.168.4.191相对应）
2）打开两台squid的日志，用tail -f的方式，然后访问 www.jialisong.com，可以看到两边的日志的内容是不一样的，从其中一台squid的日志中找一条静态文件的url，
通过ie 打开，可以发现，这个url的访问只投到同一台squid上，如此实现了，squid缓存的内容不同，从而增加了squid的缓存容量
当然，如果中间只使用一台nginx，自然这台会成为单点而造成隐患，所以可是使用heartbeat做成双机，增加系统的可靠性。