haproxy+varnish+LNMP+memcached配置

实验：

Haproxy服务器： 192.168.43.129 安装haproxy

varnish服务器：192.168.43.130 安装 varnish

web服务器：    192.168.43.128 安装 nginx php-cgi  mysql memcache memadmin discuz

首先做好hosts文件的解析 在192.168.43.130 hosts中添加

192.168.43.128  www.kerry.com 

192.168.43.128  kerry.com

一 web服务安装

1.1  在192.168.43.128中安装lnmp环境 这个在论坛很多 安装完毕之后启动服务 略过安装

解压缩discuz压缩包，将upload拷贝到web目录下进行discuz安装 这个略过

1.2 安装memecache

#wget  http://memcached.googlecode.com/files/memcached-1.4.13.tar.gz  

#wget  http://www.monkey.org/~provos/libevent-1.3.tar.gz  

1.3 先安装libevent

# tar zxvf libevent-1.3.tar.gz 

# cd libevent-1.3 

# ./configure --prefix=/usr 

# make 

# make instal

1.4 安装memcached，同时需要安装中指定libevent的安装位置：

# cd /tmp 

# tar zxvf memcached-1.4.13.tar.gz 

# cd memcached-1.4.13 

# ./configure --with-libevent=/usr 

# make 

# make install 

安装完成后会把memcached放到 /usr/local/bin/memcached 

1.5 fastcgi 安装memcache扩展模块，让php支持memcached

在http://pecl.php.net/package/memcache 选择相应想要下载的memcache版本

安装PHP的memcache扩展

#tar vxzf memcache-2.2.1.tgz 

#cd memcache-2.2.1 

#/usr/local/webserver/php/bin/phpize 

#./configure--enable-memcache --with-php-config=/usr/local/webserver/php/bin/php-config --with-zlib-dir 

#make 

#make install 

上述安装完后会有类似这样的提示

Installing shared extensions: /usr/local/webserver/php/lib/php/extensions/no-debug-non-zts-20060613/  ################即为安装成功

1.6 在php.ini 添加一行来载入memcache扩展：

extension = "memcache.so" 

1.7 启动memecache

# /usr/local/bin/memcached -d -m 1024 -u root -l 127.0.0.1 -p 11211 -c 10240 -P /tmp/memcached.pid 

    netstat -antp|grep memcache 检查端口

    

memcache启动参数备注： 
-d选项是启动一个守护进程， 
-m是分配给Memcache使用的内存数量，单位是MB，我这里是10MB， 
-u是运行Memcache的用户，我这里是root， 
-l是监听的服务器IP地址，为了安全我这里指定了 127.0.0.1， 
-p是设置Memcache监听的端口，我这里设置了11211，最好是1024以上的端口， 
-c选项是最大运行的并发连接数，默认是1024，我这里设置10240，按照你服务器的负载量来设定 
-P是设置保存Memcache的pid文件，我这里是保存在 /tmp/memcached.pid，

Memcached安装完毕。我们来测试一下是否能缓存mysql数据。我们在此服务器上架设一个php网站，此处我先用的是占用资源比较大的magento程序网站，调试好配置文件，绑定一下host文件。看网站是否能正常打开

所以我就安装一个工具memadmin 

#wget http://www.junopen.com/memadmin/memadmin-1.0.12.tar.gz

解压缩放到web的目录下 直接kerry.com/memadmin 就可以打开 默认账号密码为admin/admin。可以进去看看memcache的命中率了

##################以上web服务器就搭建完成了，现在开始搭建varnish服务器##############################

二 varnish服务器安装

groupadd www

useradd www -g www -s /sbin/nologin

mkdir -p /data/varnish/{cache,logs}

chmod +w /data/varnish/{cache,logs}

chown -R www:www /data/varnish/{cache,logs}

cd /soft

yum install -y automake autoconf libtool ncurses-devel libxslt groff pcre-devel pkgconfig 

wget http://sourceforge.net/projects/varnish/files/varnish/2.1.3/varnish-2.1.3.tar.gz/download

tar -zxvf varnish-2.1.3.tar.gz

cd varnish-2.1.3

./configure --prefix=/usr/local/varnish

make;make install

2.1 配置文件修改   

vim /usr/local/varnish/etc/varnish/kerry.vcl

backend kerry {              #定义后端服务器名

 .host = "192.168.43.128";    #定义后端服务器IP

 .port = "80";      #定义后端服务器端口

}

backend king {

 .host = "192.168.43.131";

 .port = "80";

}

#定义访问控制列表,充许那些IP清除varnish 缓存

acl local {

 "localhost";

 "127.0.0.1";

}

#判断host请求针对那个后端服务器

sub vcl_recv {

 if (req.http.host ~ "^(www.)?kerry.com$") {  #泛域名的写法"^(.*.)?kerry.com$"

  set req.backend = kerry;

 }

 elsif (req.http.host ~ "^(www.)?king.com$") {

  set req.backend = king;

 }

 else {

  error 404 "Unknown HostName!"; #如果都不匹配，返回404错误

 } 

 #不充许非访问控制列表的IP进行varnish缓存清除

 if(req.request == "PURGE") {

  if (!client.ip ~ local) {

   error 405 "Not Allowed.";

   return (lookup);

   }

 }

 #清除url中有jpg|png|gif等文件的cookie

 if (req.request == "GET" && req.url ~ "\.(jpg|png|gif|swf|jpeg|ico)$") {

  unset req.http.cookie;

 }

 #取消服务器上p_w_picpaths目录下所有文件的cookie

 if (req.url ~ "^/p_w_picpaths") {

  unset req.http.cookie;

 }

 #判断req.http.X-Forwarded-For,如果前端有多重反向代理,这样可以获取客户端IP地址。

 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") {

  return (pipe);

 }

 #针对请求和url地址判断，是否在varnish缓存里查找

 if (req.request != "GET" && req.request != "HEAD") {

  return (pass);

 } ## 对非GET|HEAD请求的直接转发给后端服务器

 if (req.http.Authorization || req.http.Cookie) {

  return (pass);

 }

 if (req.request == "GET" && req.url ~ "\.(php)($|\?)") {

  return (pass);

  } #对GET请求，且url里以.php和.php?结尾的，直接转发给后端服务器

      return (lookup);

 }  #除了以上的访问以外，都在varnish缓存里查找

sub vcl_pipe {

 return (pipe);

}

sub vcl_pass {

 return (pass);

}

sub vcl_hash {

 set req.hash += req.url;

 if (req.http.host) {

  set req.hash += req.http.host;

 } else {

  set req.hash += server.ip;

 }

 return (hash);

}

sub vcl_hit {

 if (!obj.cacheable) {

  return (pass);

 }

 if (req.request == "PURGE") {

         set obj.ttl = 0s;

         error 200 "Purged.";

      }

 return (deliver);

}

sub vcl_miss {

 return (fetch);

}

sub vcl_fetch {

 if (!beresp.cacheable) {

  return (pass);

 }

 if (beresp.http.Set-Cookie) {

  return (pass);

 }

 #WEB服务器指明不缓存的内容，varnish服务器不缓存

 if (beresp.http.Pragma ~ "no-cache" ||

     beresp.http.Cache-Control ~ "no-cache" ||

            beresp.http.Cache-Control ~ "private") {

  return (pass);

      }

      #对.txt .js .shtml结尾的URL缓存时间设置1小时，对其他的URL缓存时间设置为10天

 if (req.request == "GET" && req.url ~ "\.(txt|js|css|shtml|html|htm)$") {

               set beresp.ttl = 3600s;

 }

 else {

               set beresp.ttl = 10d;

 }

 return (deliver);

}

#添加在页面head头信息中查看缓存命中情况

sub vcl_deliver {

 set resp.http.x-hits = obj.hits ;

 if (obj.hits > 0) {

  set resp.http.X-Cache = "HIT cqtel-bbs";

 }

 else {

      set resp.http.X-Cache = "MISS cqtel-bbs";

 }

}

sub vcl_error {

 set obj.http.Content-Type = "text/html; charset=utf-8";

 synthetic {"

   

     

   

   

     

Error "} obj.status " " obj.response {"

     

"} obj.response {"

     

Guru Meditation:

     

XID: "} req.xid {"

     

---

     

         bbs cache server

     

   

"};

 return (deliver);

}

##########################################################################################

############配置文件的解析详情参考http://ixdba.blog.51cto.com/2895551/682555##############

##########################################################################################

2.2 ln /usr/local/varnish/var/varnish/clinet1/_.vsl  /data/varnish/cache/_.vsl

2.3 启动服务

/usr/local/varnish/sbin/varnishd -u www -g www -f /usr/local/varnish/etc/varnish/kerry.vcl -a 192.168.43.130:80 -s file,/data/varnish/cache/varnish_cache.data,1G -w 1024,51200,10 -t 3600 -T 192.168.43.130:3000

echo "/usr/local/varnish/sbin/varnishd -u www -g www -f /usr/local/varnish/etc/varnish/kerry.vcl -a 192.168.43.130:80 -s file,/data/varnish/cache/varnish_cache.data,1G -w 1024,51200,10 -t 3600 -T 192.168.43.130:3000"  >> /etc/rc.local

1. 参数: 
2. -u 以什么用运行 
3. -g 以什么组运行 
4. -f varnish配置文件 
5. -a 绑定IP和端口 
6. -s varnish缓存文件位置与大小 
7. -w 最小，最大线程和超时时间 
8. -T varnish管理端口，主要用来清除缓存
9. -p client_http11=on 支持http1.1协议
10. -P(大P) /usr/local/varnish/var/varnish.pid 指定其进程码文件的位置，实现管理
11. 停止varnish
12. pkill varnishd  #结束varnishd进程

2.4 /usr/local/varnish/bin/varnishncsa -w /data/varnish/logs/varnish.log &

echo "/usr/local/varnish/bin/varnishncsa -w /data/varnish/logs/varnish.log &" >> /etc/rc.local

参数: -w 指定varnish访问日志要写入的目录与文件

2.5 vi /root/cut_varnish_log.sh

1. #!/bin/sh
2. logs_path=/data/varnish/logs
3. vlog=${logs_path}/varnish.log
4. date=$(date -d "yesterday" +"%Y-%m-%d")
5. pkill -9 varnishncsa
6. mkdir -p ${logs_path}/$(date -d "yesterday" +"%Y")/$(date -d "yesterday" +"%m")/
7. mv /data/varnish/logs/varnish.log ${logs_path}/$(date -d "yesterday" +"%Y")/$(date -d "yesterday" +"%m")/varnish-${date}.log
8. /usr/local/varnish/bin/varnishncsa -w /data/varnish/logs/varnish.log &

使用计划任务,每天晚上凌晨00点运行日志切割脚本

echo "0 0 * * * /root/cut_varnish_log.sh" >> /etc/crontab

2.6 查看命中率

/usr/local/varnish/bin/varnishstat –n /data/varnish/cache/varnish_cache.data

三 varnish知识补全

3.1  ======================Varnish 缓存清除 ======================

/usr/local/varnish/bin/varnishadm -T 192.168.43.130:3000 purge "req.http.host ~ www.kerry.com$ && req.url ~ /static/p_w_picpath/tp.php"

说明： 

192.168.9.201:3000 为被清除缓存服务器地址 

www.kerry.com 为被清除的域名 

/static/p_w_picpath/tp.php 为被清除的url地址列表

清除所有缓存

/usr/local/varnish/bin/varnishadm -T 192.168.43.130:3000 url.purge *$

清除p_w_picpath目录下所有缓存 

/usr/local/varnish/bin/varnishadm -T 192.168.43.130:3000 url.purge /p_w_picpath/

3.2 ======================内核优化  ======================

vi /etc/sysctl.conf

net.ipv4.tcp_syncookies = 1

net.ipv4.tcp_tw_reuse = 1

net.ipv4.tcp_tw_recycle = 1

#net.ipv4.tcp_fin_timeout = 30

#net.ipv4.tcp_keepalive_time = 300

net.ipv4.ip_local_port_range = 1024 65000

net.ipv4.tcp_max_syn_backlog = 8192

net.ipv4.tcp_max_tw_buckets = 5000

net.ipv4.tcp_max_syn_backlog = 65536

net.core.netdev_max_backlog =  32768

net.core.somaxconn = 32768

net.core.wmem_default = 8388608

net.core.rmem_default = 8388608

net.core.rmem_max = 16777216

net.core.wmem_max = 16777216

net.ipv4.tcp_timestamps = 0

net.ipv4.tcp_synack_retries = 2

net.ipv4.tcp_syn_retries = 2

net.ipv4.tcp_tw_recycle = 1

#net.ipv4.tcp_tw_len = 1

net.ipv4.tcp_tw_reuse = 1

net.ipv4.tcp_mem = 94500000 915000000 927000000

net.ipv4.tcp_max_orphans = 3276800

/sbin/sysctl -p

3.3 ===================== Varnish添加到服务自启动 ======================

vi /etc/init.d/varnish

#! /bin/sh

#

# varnish Control the varnish HTTP accelerator

#

# chkconfig: - 90 10

# description: Varnish is a high-perfomance HTTP accelerator

# processname: varnishd

# config: /etc/sysconfig/varnish

# pidfile: /var/run/varnish/varnishd.pid

### BEGIN INIT INFO

# Provides: varnish

# Required-Start: $network $local_fs $remote_fs

# Required-Stop: $network $local_fs $remote_fs

# Should-Start: $syslog

# Short-Description: start and stop varnishd

# Description: Varnish is a high-perfomance HTTP accelerator

### END INIT INFO

# Source function library.

. /etc/init.d/functions

retval=0

pidfile=/var/run/varnish.pid

exec="/usr/local/varnish/sbin/varnishd"

prog="varnishd"

config="/usr/local/varnish/etc/varnish/varnish"

lockfile="/var/lock/subsys/varnish"

# Include varnish defaults

[ -e /usr/local/varnish/etc/varnish/varnish ] && . /usr/local/varnish/etc/varnish/varnish

start() {

    if [ ! -x $exec ]

    then

        echo $exec not found

        exit 5

    fi

    if [ ! -f $config ]

    then

        echo $config not found

        exit 6

    fi

    echo -n "Starting varnish HTTP accelerator: "

    # Open files (usually 1024, which is way too small for varnish)

    ulimit -n ${NFILES:-131072}

    # Varnish wants to lock shared memory log in memory. 

    ulimit -l ${MEMLOCK:-82000}

        # $DAEMON_OPTS is set in /etc/sysconfig/varnish. At least, one

        # has to set up a backend, or /tmp will be used, which is a bad idea.

    if [ "$DAEMON_OPTS" = "" ]; then

        echo "\$DAEMON_OPTS empty."

        echo -n "Please put configuration options in $config"

        return 6

    else

        # Varnish always gives output on STDOUT

        daemon   $exec -P $pidfile "$DAEMON_OPTS" > /dev/null 2>&1

        retval=$?

        if [ $retval -eq 0 ]

        then

            touch $lockfile

            echo_success

            echo

        else

            echo_failure

        fi

        return $retval

    fi

}

stop() {

    echo -n "Stopping varnish HTTP accelerator: "

    killproc $prog

    retval=$?

    echo

    [ $retval -eq 0 ] && rm -f $lockfile

    return $retval

}

restart() {

    stop

    start

}

reload() {

    restart

}

force_reload() {

    restart

}

rh_status() {

    status $prog

}

rh_status_q() {

    rh_status >/dev/null 2>&1

}

# See how we were called.

case "$1" in

    start)

        rh_status_q && exit 0

        $1

        ;;

    stop)

        rh_status_q || exit 0

        $1

        ;;

    restart)

        $1

        ;;

    reload)

        rh_status_q || exit 7

        $1

        ;;

    force-reload)

        force_reload

        ;;

    status)

        rh_status

        ;;

    condrestart|try-restart)

        rh_status_q || exit 0

        restart

        ;;

    *)

    echo "Usage: $0 {start|stop|status|restart|condrestart|try-restart|reload|force-reload}"

    exit 2

esac

exit $?

varnish的配置调用文件,是用来告诉程序从哪里读取配置文件,启动参数有哪些等

vim /usr/local/varnish/etc/varnish/varnish

# Configuration file for varnish

#

# Configuration file for varnish

#

# Configuration file for varnish

#

# /etc/init.d/varnish expects the variable $DAEMON_OPTS to be set from this

# shell script fragment.

#

# Maximum number of open files (for ulimit -n)

NFILES=131072

# Locked shared memory (for ulimit -l)

# Default log size is 82MB + header

MEMLOCK=1000000

## Alternative 2, Configuration with VCL

DAEMON_OPTS="-a 192.168.43.130:80 \

             -f /usr/local/varnish/etc/varnish/kerry.vcl \

             -T 192.168.43.130:3000 \

             -u www -g www \

             -n /data/varnish/cache \

             -s file,/data/varnish/cache/varnish_cache.data,1G"

             

##### /data/varnish/cache/varnish_cache.data,1G这个设置在32位的系统不能超过2G

3.4 添加到系统服务，开机自启动

chmod +x /etc/init.d/varnish 

/sbin/chkconfig --add varnish

/sbin/chkconfig --level 2345 varnish on            

四 haproxy安装

4.1 将本机的域名解析修改为 

192.168.43.129 www.kerry.com

192.168.43.129 kerry.com

4.2 下载安装

wget http://haproxy.1wt.eu/download/1.4/src/haproxy-1.4.19.tar.gz
#tar -xvf haproxy-1.4.19.tar.gz
#cd haproxy-1.4.19
#make TARGET=linux26 PREFIX=/usr/local/haproxy
#make install PREFIX=/usr/local/haproxy

4.3 配置文件修改

# cd /usr/local/haproxy
# vim haproxy.conf
global
        maxconn 5120
        chroot /usr/local/haproxy
        uid 99
        gid 99
        daemon
        quiet
        nbproc  2
        pidfile /usr/local/haproxy/haproxy.pid
defaults
        log     global
        mode    http
        option  httplog
        option  dontlognull
        log 127.0.0.1 local3
        retries 3
        option redispatch
        maxconn 20000
        contimeout      5000
        clitimeout      50000
        srvtimeout      50000
listen www.kerry.com 0.0.0.0:80
       mode http
       stats uri /status                     (后端服务器状态查看地址)
       stats realm Haproxy\ statistics
       stats auth admin:admin              （状态查看页面登陆帐号密码）
       balance source    （调度算法，调度算法有很多，我这里用source是和nginx的ip_hash同理，解决session问题)
       option httpclose
       option forwardfor
server app1_1 192.168.43.130:80 cookie app1inst1 check inter 2000 rise 2 fall 5
## server app1_2 192.168.1.21:80 cookie app1inst2 check inter 2000 rise 2 fall 5
###这个地方时你有几台varnish服务器就添加多少 每台一行

4.4 启动服务 ./sbin/haproxy  -f haproxy.conf 

[root@clinet2 haproxy]# netstat -antp|grep haproxy

tcp        0      0 0.0.0.0:80                  0.0.0.0:*                   LISTEN      1711/./sbin/haproxy 

4.5 查看haproxy 状态

http://www.kerry.com/status