安裝配置Varnish3.0手記


 
在內網的vps上安裝Varnish的最新版3.0來用做測試

Varnish有centos下的rpm和ubuntu下的deb二進制包下載,但我選擇的是源碼編譯安裝

先下載 wget http://repo.varnish-cache.org/source/varnish-3.0.0.tar.gz

tar -xzvf varnish-3.0.0.tar.gz

 

yum install gcc

yum install gcc-c++ libstdc++-devel

yum install -y httpd-devel pcre perl pcre-devel zlib zlib-devel GeoIP GeoIP-devel

 

Varnish需要pcre支持,如果服務器沒有安裝pcre-devel會提示需要指定prce目錄。
Centos服務器上使用yum install pcre-devel安裝

 

解壓varnish進行編譯安裝

cd varnish-3.0.0

./autogen.sh

./configure --prefix=/usr/local/varnish PKG_CONFIG_PATH=/usr/lib/pkgconfig
 

make

make install

安裝成功后,安裝的目錄是

cd /usr/local/varnish/sbin

./varnishd -V

查看版本,看是否安裝成功

測試varnish

先編輯配置文件

vi /usr/local/varnish/etc/varnish/default.vcl

 

 

# This  is a basic VCL configuration file  for varnish.  See the vcl( 7)
# man page  for details on VCL syntax and semantics.
#
# Default backend definition.  Set  this to point to your content
# server.
#
 backend  default {
     .host =  " 192.168.88.156 ";
     .port =  " 80 ";
     ###下面三行為新加配
     .connect_timeout = 1s;
     .first_byte_timeout = 5s;
     .between_bytes_timeout = 2s;
 }
#
# Below  is a commented- out copy of the  default VCL logic.  If you
# redefine any of these subroutines, the built- in logic will be
# appended to your code.
 sub vcl_recv {
      if (req.restarts ==  0) {
         if (req.http.x-forwarded- for) {
             set req.http.X-Forwarded-For =
                req.http.X-Forwarded-For +  " " + client.ip;
        }  else {
             set req.http.X-Forwarded-For = client.ip;
        }
     }
      if (req.request !=  " GET " &&
       req.request !=  " HEAD " &&
       req.request !=  " PUT " &&
       req.request !=  " POST " &&
       req.request !=  " TRACE " &&
       req.request !=  " OPTIONS " &&
       req.request !=  " DELETE ") {
          /*  Non-RFC2616 or CONNECT which is weird.  */
          return (pipe);
     }
      if (req.request !=  " GET " && req.request !=  " HEAD ") {
          /*  We only deal with GET and HEAD by default  */
          return (pass);
     }
      if (req.http.Authorization || req.http.Cookie) {
          /*  Not cacheable by default  */
          return (pass);
     }
      return (lookup);
 }
#
 sub vcl_pipe {
#     # Note that only the first request to the backend will have
#     # X-Forwarded-For  set.  If you use X-Forwarded-For and want to
#     # have it  set  for all requests, make sure to have:
#     #  set bereq.http.connection =  " close ";
#     # here.  It  is not  set by  default  as it might  break some broken web
#     # applications, like IIS with NTLM authentication.
      return (pipe);
 }
#
 sub vcl_pass {
      return (pass);
 }
#
 sub vcl_hash {
     hash_data(req.url);
      if (req.http.host) {
         hash_data(req.http.host);
     }  else {
         hash_data(server.ip);
     }
      return (hash);
 }
#
 sub vcl_hit {
      return (deliver);
 }
#
 sub vcl_miss {
      return (fetch);
 }
#
 sub vcl_fetch {
      if (beresp.ttl <= 0s ||
         beresp.http.Set-Cookie ||
         beresp.http.Vary ==  " * ") {
                 /*
                 * Mark as "Hit-For-Pass" for the next 2 minutes
                 
*/
                 set beresp.ttl =  120 s;
                 return (hit_for_pass);
     }
      return (deliver);
 }
#
 sub vcl_deliver {
      return (deliver);
 }
#
# sub vcl_error {
#      set obj.http.Content-Type =  " text/html; charset=utf-8 ";
#      set obj.http.Retry-After =  " 5 ";
#     synthetic { "
# <?xml version= " 1.0 " encoding= " utf-8 "?>
# <!DOCTYPE html PUBLIC  " -//W3C//DTD XHTML 1.0 Strict//EN "
#   " http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd ">
# <html>
#   <head>
#     <title> " } + obj.status +  "  "  + obj.response + { "</title>
#   </head>
#   <body>
#     <h1>Error  " } + obj.status +  "  "  + obj.response + { "</h1>
#     <p> " } + obj.response + { "</p>
#     <h3>Guru Meditation:</h3>
#     <p>XID:  " } + req.xid + { "</p>
#     <hr>
#     <p>Varnish cache server</p>
#   </body>
# </html>
" };
#      return (deliver);
# }
#
 sub vcl_init {
         return (ok);
 }
#
 sub vcl_fini {
         return (ok);
 }

 

/usr/local/varnish/sbin/varnishd -f /usr/local/varnish/etc/varnish/default.vcl -s malloc,1024m -T 127.0.0.1:200 -a 0.0.0.0:80

 

現在可以打開varnish了

http://192.168.88.1/    (開啟varnish的端口號)

戴圖顯示,nginx 為177上運行的web服務器,出現了varnish字樣,表示緩存架設成功,已正常工作了。

 

啟動varnishncsa用來將Varnish訪問日志寫入日志文件:

/usr/local/varnish/bin/varnishncsa -n /var/vcache -w /var/logs/varnish.log &



配置開機自動啟動Varnish

vi /etc/rc.local


  在末尾增加以下內容:

引用
ulimit -SHn 51200
/usr/local/varnish/sbin/varnishd -n /var/vcache -f /usr/local/varnish/vcl.conf -a 0.0.0.0:80 -s file,/var/vcache/varnish_cache.data,1G -g www -u www -w 30000,51200,10 -T 127.0.0.1:3500 -p client_http11=on
/usr/local/varnish/bin/varnishncsa -n /var/vcache -w /var/logs/youvideo.log &



 優化Linux內核參數

vi /etc/sysctl.conf


  在末尾增加以下內容:

引用
net.ipv4.tcp_fin_timeout = 30
net.ipv4.tcp_keepalive_time = 300
net.ipv4.tcp_syncookies = 1
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_tw_recycle = 1
net.ipv4.ip_local_port_range = 5000    65000


 查看Varnish服務器連接數與命中率
/usr/local/varnish/bin/varnishstat

 

配置啟動腳本: 

mkdir /cache/varnish/V -p

vi /usr/local/varnish/sbin/start.sh

 

#!/bin/sh
#file:start.sh
date -u
/usr/local/varnish/sbin/varnishd -a 0.0.0.0:80  -s file,/cache/varnish/V,1024m  -f         /usr/local/varnish/etc/varnish/default.vcl  -p thread_pool_max=1500  -p thread_pools=5  -p listen_depth=512 

chmod 777 start.sh

運行:

/usr/local/varnish/sbin/start.sh


免責聲明!

本站轉載的文章為個人學習借鑒使用,本站對版權不負任何法律責任。如果侵犯了您的隱私權益,請聯系本站郵箱yoyou2525@163.com刪除。



 
粵ICP備18138465號   © 2018-2025 CODEPRJ.COM