当前位置:技术首页 >> 新闻动态 >> 业界资讯 >> Nginx 0.7.x + PHP 5.2.6(FastCGI)建胜过Apache十倍的Web服务器
Nginx 0.7.x + PHP 5.2.6(FastCGI)建胜过Apache十倍的Web服务器
2008-10-23 01:00:49  作者:  来源:互联网  浏览次数:9  文字大小:【】【】【
  •    Nginx ("engine x") 是一个高性能的 HTTP 和反向代理服务器,也是一个IMAP/POP3/SMTP 代理服务器。 Nginx 是由 Igor Sysoev为俄罗斯访问量第二的 Rambler.ru 站点开发的,它已经在该站点运行超过两年半了。Igor 将源代码以类BSD许可证的形式发布。

    Nginx ("engine x") 是一个高性能的 HTTP 和反向代理服务器,也是一个IMAP/POP3/SMTP 代理服务器。 Nginx 是由 Igor Sysoev为俄罗斯访问量第二的 Rambler.ru 站点开发的,它已经在该站点运行超过两年半了。Igor 将源代码以类BSD许可证的形式发布。
    Nginx 超越 Apache 的高性能和稳定性,使得国内使用 Nginx 作为 Web 服务器的网站也越来越多,其中包括新浪博客、新浪播客、网易新闻等门户网站频道,六间房、56.com等视频分享网站,Discuz!官方论坛、水木社区等知名论坛,豆瓣、YUPOO相册、海内SNS、迅雷在线等新兴Web 2.0网站。
    Nginx 的官方中文维基:http://wiki.codemongers.com/NginxChs
    在高并发连接的情况下,Nginx是Apache服务器不错的替代品。Nginx同时也可以作为7层负载均衡服务器来使用。根据我的测试结果,Nginx 0.7.17 + PHP 5.2.6 (FastCGI) 可以承受3万以上的并发连接数,相当于同等环境下Apache的10倍。
    根据我的经验,4GB内存的服务器+Apache(prefork模式)一般只能处理3000个并发连接,因为它们将占用3GB以上的内存,还得为系统预留1GB的内存。我曾经就有两台Apache服务器,因为在配置文件中设置的MaxClients为4000,当Apache并发连接数达到3800 时,导致服务器内存和Swap空间用满而崩溃。
    而这台 Nginx 0.7.17 + PHP 5.2.6 (FastCGI) 服务器在3万并发连接下,开启的10个Nginx进程消耗150M内存(15M*10=150M),开启的64个php-cgi进程消耗1280M内存(20M*64=1280M),加上系统自身消耗的内存,总共消耗不到2GB内存。如果服务器内存较小,完全可以只开启25个php-cgi进程,这样 php-cgi消耗的总内存数才500M。
    在3万并发连接下,访问Nginx 0.7.17 + PHP 5.2.6 (FastCGI) 服务器的PHP程序,仍然速度飞快。下图为Nginx的状态监控页面,显示的活动连接数为28457(关于Nginx的监控页配置,会在本文接下来所给出的Nginx配置文件中写明):

    我生产环境下的两台Nginx + PHP5(FastCGI)服务器,跑多个一般复杂的纯PHP动态程序,单台Nginx + PHP5(FastCGI)服务器跑PHP动态程序的处理能力已经超过“700次请求/秒”,相当于每天可以承受6000万(700*60*60*24=60480000)的访问量(更多信息见此),而服务器的系统负载也不高:

安装步骤:
    (系统要求:Linux 2.6+ 内核,本文中的Linux操作系统为CentOS 5.1,另在RedHat AS4上也安装成功)

获取相关开源程序:

    利用CentOS Linux系统自带的yum命令安装、升级所需的程序库(RedHat等其他Linux发行版可从安装光盘中找到这些程序库的RPM包,进行安装):
    sudo -s
    LANG=C
    yum -y install gcc gcc-c++ autoconf libjpeg libjpeg-devel libpng libpng-devel
    freetype freetype-devel libxml2 libxml2-devel zlib zlib-devel glibc glibc-devel
    glib2 glib2-devel bzip2 bzip2-devel ncurses ncurses-devel curl curl-devel
    RedHat等其他Linux发行版可从安装光盘中找到这些程序库的RPM包(事先可通过类似“rpm -qa | grep libjpeg”的命令查看所需的RPM包是否存在,通常是“xxx-devel”不存在,需要安装)。RedHat可以直接利用CentOS的RPM包安装,以下是RPM包下载网址:
 1.RedHat AS4 & CentOS 4
  http://mirror.be10.com/centos/4/os/i386/CentOS/RPMS/
  http://mirror.be10.com/centos/4/os/x86_64/CentOS/RPMS/
 2.RedHat AS5 & CentOS 5
  http://mirror.be10.com/centos/5/os/i386/CentOS/
  http://mirror.be10.com/centos/5/os/x86_64/CentOS/
 3.RPM包搜索网站
  http://rpm.pbone.net/
  http://www.rpmfind.net/

下载程序源码包:
    本文中提到的所有开源软件为截止到2008年8月28日的最新稳定版。
    mkdir -p /data0/software
    cd /data0/software
    wget http://sysoev.ru/nginx/nginx-0.7.17.tar.gz
    wget http://www.php.net/get/php-5.2.6.tar.gz/from/this/mirror
    wget http://php-fpm.anight.org/downloads/head/php-5.2.6-fpm-0.5.9.diff.gz
    wget http://dev.mysql.com/get/Downloads/MySQL-5.1/mysql-5.1.26-rc.tar.gz/from/http://mirror.x10.com/mirror/mysql/
    wget http://blog.s135.com/soft/linux/mysql/mysql-5.1.26-rc.tar.gz
    wget http://ftp.gnu.org/pub/gnu/libiconv/libiconv-1.12.tar.gz
    wget "http://downloads.sourceforge.net/mcrypt/libmcrypt-2.5.8.tar.gz?modtime=1171868460&big_mirror=0"
    wget http://mirror.optus.net/sourceforge/m/mc/mcrypt/libmcrypt-2.5.8.tar.gz
    wget "http://downloads.sourceforge.net/mcrypt/mcrypt-2.6.7.tar.gz?modtime=1194463373&big_mirror=0"
    wget http://mirror.optus.net/sourceforge/m/mc/mcrypt/mcrypt-2.6.7.tar.gz
    wget http://pecl.php.net/get/memcache-2.2.3.tgz
    wget "http://downloads.sourceforge.net/mhash/mhash-0.9.9.tar.gz?modtime=1175740843&big_mirror=0"
    wget http://mirror.optus.net/sourceforge/m/mh/mhash/mhash-0.9.9.tar.gz
    wget ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre/pcre-7.7.tar.gz
    wget http://bart.eaccelerator.net/source/0.9.5.3/eaccelerator-0.9.5.3.tar.bz2

安装PHP 5.2.6(FastCGI模式)
    1 编译安装PHP 5.2.6所需的支持库:
    tar zxvf libiconv-1.12.tar.gz
    cd libiconv-1.12/
    ./configure --prefix=/usr/local
    make
    make install
    cd ../
    tar zxvf libmcrypt-2.5.8.tar.gz
    cd libmcrypt-2.5.8/
    ./configure
    make
    make install
    /sbin/ldconfig
    cd libltdl/
    ./configure --enable-ltdl-install
    make
    make install
    cd ../../
    tar zxvf mhash-0.9.9.tar.gz
    cd mhash-0.9.9/
    ./configure
    make
    make install
    cd ../
    cp /usr/local/lib/libmcrypt.* /usr/lib
    ln -s /usr/local/lib/libmhash.so.2 /usr/lib/libmhash.so.2
    tar zxvf mcrypt-2.6.7.tar.gz
    cd mcrypt-2.6.7/
    ./configure
    make
    make install
    cd ../

2 编译安装MySQL 5.1.26-rc
    /usr/sbin/groupadd mysql
    /usr/sbin/useradd -g mysql mysql
    tar zxvf mysql-5.1.26-rc.tar.gz
    cd mysql-5.1.26-rc/
    ./configure --prefix=/usr/local/webserver/mysql/ --enable-assembler --with-extra-charsets=complex --enable-thread-safe-client --with-big-tables --with-readline --with-ssl --with-embedded-server --enable-local-infile
    make && make install
    chmod +w /usr/local/webserver/mysql
    chown -R mysql:mysql /usr/local/webserver/mysql
    cp support-files/my-medium.cnf /usr/local/webserver/mysql/my.cnf
    cd ../
    附:以下为附加步骤,如果你想在这台服务器上运行MySQL数据库,则执行以下两步。如果你只是希望让PHP支持MySQL扩展库,能够连接其他服务器上的MySQL数据库,那么,以下两步无需执行。

3 以mysql用户帐号的身份建立数据表:
    /usr/local/webserver/mysql/bin/mysql_install_db --defaults-file=/usr/local/webserver/mysql/my.cnf --asedir=/usr/local/webserver/mysql --datadir=/usr/local/webserver/mysql/data --user=mysql --pid-file=/usr/local/webserver/mysql/mysql.pid --skip-locking --port=3306 --socket=/tmp/mysql.sock

4 启动MySQL(最后的&表示在后台运行)
    /bin/sh /usr/local/webserver/mysql/bin/mysqld_safe --defaults-file=/usr/local/webserver/mysql/my.cnf &

5 编译安装PHP(FastCGI模式)
    tar zxvf php-5.2.6.tar.gz
    gzip -cd php-5.2.6-fpm-0.5.9.diff.gz | patch -d php-5.2.6 -p1
    cd php-5.2.6/
    ./configure --prefix=/usr/local/webserver/php --with-config-file-path=/usr/local/webserver/php/etc --with-mysql=/usr/local/webserver/mysql --with-mysqli=/usr/local/webserver/mysql/bin/mysql_config --with-iconv-dir=/usr/local --with-freetype-dir --with-jpeg-dir --with-png-dir --with-zlib --with-libxml-dir=/usr --enable-xml --disable-rpath --enable-discard-path
--enable-safe-mode --enable-bcmath --enable-shmop --enable-sysvsem --enable-inline-optimization --with-curl --with-curlwrappers --enable-mbregex --enable-fastcgi --enable-fpm --enable-force-cgi-redirect --enable-mbstring --with-mcrypt --with-gd --enable-gd-native-ttf --with-openssl
    sed -i 's#-lz -lm -lxml2 -lz -lm -lxml2 -lz -lm -lcrypt#& -liconv#' Makefile
    make
    make install
    cp php.ini-dist /usr/local/webserver/php/etc/php.ini
    cd ../

6 编译安装PHP5扩展模块
    tar zxvf memcache-2.2.3.tgz
    cd memcache-2.2.3/
    /usr/local/webserver/php/bin/phpize
    ./configure --with-php-config=/usr/local/webserver/php/bin/php-config
    make
    make install
    cd ../
    tar jxvf eaccelerator-0.9.5.3.tar.bz2
    cd eaccelerator-0.9.5.3/
    /usr/local/webserver/php/bin/phpize
    ./configure --enable-eaccelerator=shared --with-php-config=/usr/local/webserver/php/bin/php-config
    make
    make install
    cd ../
    修改php.ini文件
    手工修改:查找/usr/local/webserver/php/etc/php.ini中的extension_dir = "./" 修改为extension_dir = "/usr/local/webserver/php/lib/php/extensions/no-debug-non-zts-20060613/"
    并在此行后增加以下几行,然后保存:
    extension = "memcache.so"
    再查找output_buffering = Off
    修改为output_buffering = On
    自动修改:若嫌手工修改麻烦,可执行以下shell命令,自动完成对php.ini文件的修改:
    sed -i 's#extension_dir = "./"#extension_dir = "/usr/local/webserver/php/lib/php/extensions/no-debug-non-zts-20060613/"\nextension = "memcache.so"\n#'
    /usr/local/webserver/php/etc/php.ini
    sed -i 's#output_buffering = Off#output_buffering = On#'
    /usr/local/webserver/php/etc/php.ini

8 配置eAccelerator加速PHP:
    mkdir -p /usr/local/webserver/eaccelerator_cache
    vi /usr/local/webserver/php/etc/php.ini
    按shift+g键跳到配置文件的最末尾,加上以下配置信息:
    [eaccelerator]
    zend_extension="/usr/local/webserver/php/lib/php/extensions/no-debug-non-zts-20060613/eaccelerator.so"
    eaccelerator.shm_size="128"
    eaccelerator.cache_dir="/usr/local/webserver/eaccelerator_cache"
    eaccelerator.enable="1"
    eaccelerator.optimizer="1"
    eaccelerator.check_mtime="1"
    eaccelerator.debug="0"
    eaccelerator.filter=""
    eaccelerator.shm_max="0"
    eaccelerator.shm_ttl="300"
    eaccelerator.shm_prune_period="120"
    eaccelerator.shm_only="0"
    eaccelerator.compress="1"
    eaccelerator.compress_level="9"
    修改配置文件:
    vi /etc/sysctl.conf
    输入以下内容:
    kernel.shmmax = 134217728
    然后执行以下命令使配置生效:
    /sbin/sysctl -p

9 创建www用户和组,以及供blog.xxx.com和www.xxx.com两个虚拟主机使用的目录:
    /usr/sbin/groupadd www -g 48
    /usr/sbin/useradd -u 48 -g www www
    mkdir -p /data0/htdocs/blog
    chmod +w /data0/htdocs/blog
    chown -R www:www /data0/htdocs/blog
    mkdir -p /data0/htdocs/www
    chmod +w /data0/htdocs/www
    chown -R www:www /data0/htdocs/www

10 创建php-fpm配置文件(php-fpm是为PHP打的一个FastCGI管理补丁,可以平滑变更php.ini配置而无需重启php-cgi):
    在/usr/local/webserver/php/etc/目录中创建php-fpm.conf文件:
    rm -f /usr/local/webserver/php/etc/php-fpm.conf
    vi /usr/local/webserver/php/etc/php-fpm.conf
    输入以下内容(如果您安装 Nginx + PHP 用于程序调试,请将以下的<value name="display_errors">0</value>改为<value name="display_errors">1</value>,以便显示PHP错误信息,否则,Nginx 会报状态为500的空白错误页):
    <?xml version="1.0" ?>
    <configuration>
    All relative paths in this config are relative to php's install prefix
    <section name="global_options">
    Pid file
    <value name="pid_file">/usr/local/webserver/php/logs/php-fpm.pid</value>
    Error log file
    <value name="error_log">/usr/local/webserver/php/logs/php-fpm.log</value>
    Log level
    <value name="log_level">notice</value>
    When this amount of php processes exited with SIGSEGV or SIGBUS ...
    <value name="emergency_restart_threshold">10</value>
    ... in a less than this interval of time, a graceful restart will be initiated.
    Useful to work around accidental curruptions in accelerator's shared memory.
    <value name="emergency_restart_interval">1m</value>
    Time limit on waiting child's reaction on signals from master
    <value name="process_control_timeout">5s</value>
 
    Set to 'no' to debug fpm
    <value name="daemonize">yes</value>
    </section>
    <workers>
    <section name="pool">
    Name of pool. Used in logs and stats.
    <value name="name">default</value>
    Address to accept fastcgi requests on.
    Valid syntax is 'ip.ad.re.ss:port' or just 'port' or '/path/to/unix/socket'
    <value name="listen_address">127.0.0.1:9000</value>
    <value name="listen_options">
    Set listen(2) backlog
    <value name="backlog">-1</value>
    Set permissions for unix socket, if one used.
    In Linux read/write permissions must be set in order to allow connections from web server.
    Many BSD-derrived systems allow connections regardless of permissions.
    <value name="owner"></value>
    <value name="group"></value>
    <value name="mode">0666</value>
    </value>
    Additional php.ini defines, specific to this pool of workers.
    <value name="php_defines">
    <value name="sendmail_path">/usr/sbin/sendmail -t -i</value>
    <value name="display_errors">1</value>
    </value>
    Unix user of processes
    <value name="user">www</value>
    Unix group of processes
    <value name="group">www</value>
    Process manager settings
    <value name="pm">
    Sets style of controling worker process count.
    Valid values are 'static' and 'apache-like'
    <value name="style">static</value>
    Sets the limit on the number of simultaneous requests that will be served.
    Equivalent to Apache MaxClients directive.
    Equivalent to PHP_FCGI_CHILDREN environment in original php.fcgi Used with any pm_style.
    <value name="max_children">128</value>
    Settings group for 'apache-like' pm style
    <value name="apache_like">
    Sets the number of server processes created on startup.
    Used only when 'apache-like' pm_style is selected
    <value name="StartServers">20</value>
    Sets the desired minimum number of idle server processes.
    Used only when 'apache-like' pm_style is selected
    <value name="MinSpareServers">5</value>
    Sets the desired maximum number of idle server processes.
    Used only when 'apache-like' pm_style is selected
    <value name="MaxSpareServers">35</value>
    </value>
    </value>
    The timeout (in seconds) for serving a single request after which the worker process will be terminated Should be used when 'max_execution_time' ini option does not stop script execution for some reason '0s' means 'off'
    <value name="request_terminate_timeout">0s</value>
    The timeout (in seconds) for serving of single request after which a php backtrace will be dumped to slow.log file '0s' means 'off'
    <value name="request_slowlog_timeout">0s</value>
    The log file for slow requests
    <value name="slowlog">logs/slow.log</value>
    Set open file desc rlimit
    <value name="rlimit_files">51200</value>
    Set max core size rlimit
    <value name="rlimit_core">0</value>
    Chroot to this directory at the start, absolute path
    <value name="chroot"></value>
    Chdir to this directory at the start, absolute path
    <value name="chdir"></value>
    Redirect workers' stdout and stderr into main error log.
    If not set, they will be redirected to /dev/null, according to FastCGI specs
    <value name="catch_workers_output">yes</value>
    How much requests each process should execute before respawn.
    Useful to work around memory leaks in 3rd party libraries.
    For endless request processing please specify 0
    Equivalent to PHP_FCGI_MAX_REQUESTS
    <value name="max_requests">10240</value>
    Comma separated list of ipv4 addresses of FastCGI clients that allowed to connect.
    Equivalent to FCGI_WEB_SERVER_ADDRS environment in original php.fcgi (5.2.2+)
    Makes sense only with AF_INET listening socket.
    <value name="allowed_clients">127.0.0.1</value>
    Pass environment variables like LD_LIBRARY_PATH
    All $VARIABLEs are taken from current environment
    <value name="environment">
    <value name="HOSTNAME">$HOSTNAME</value>
    <value name="PATH">/usr/local/bin:/usr/bin:/bin</value>
    <value name="TMP">/tmp</value>
    <value name="TMPDIR">/tmp</value>
    <value name="TEMP">/tmp</value>
    <value name="OSTYPE">$OSTYPE</value>
    <value name="MACHTYPE">$MACHTYPE</value>
    <value name="MALLOC_CHECK_">2</value>
    </value>
    </section>
    </workers>
    </configuration>

11 启动php-cgi进程,监听127.0.0.1的9000端口,进程数为200(如果服务器内存小于3GB,可以只开启64个进程),用户为www:

    ulimit -SHn 51200
    /usr/local/webserver/php/sbin/php-fpm start
注:/usr/local/webserver/php/sbin/php-fpm还有其他参数,包括:start|stop|quit|restart|reload|logrotate,修改php.ini后不重启php-cgi,重新加载配置文件使用reload。

三、安装Nginx 0.7.17

1. 安装Nginx所需的pcre库:
    tar zxvf pcre-7.7.tar.gz
    cd pcre-7.7/
    ./configure
    make && make install
    cd ../

2. 安装Nginx
    tar zxvf nginx-0.7.17.tar.gz
    cd nginx-0.7.17/
    ./configure --user=www --group=www --prefix=/usr/local/webserver/nginx --with-http_stub_status_module --with-http_ssl_module
    make && make install
    cd ../

3. 创建Nginx日志目录
    mkdir -p /data1/logs
    chmod +w /data1/logs
    chown -R www:www /data1/logs

4. 创建Nginx配置文件
1. 在/usr/local/webserver/nginx/conf/目录中创建nginx.conf文件:
    rm -f /usr/local/webserver/nginx/conf/nginx.conf
    vi /usr/local/webserver/nginx/conf/nginx.conf
    输入以下内容:
 
    user  www www;
    worker_processes 8;
    error_log  /data1/logs/nginx_error.log  crit;
    pid        /usr/local/webserver/nginx/nginx.pid;
    #Specifies the value for maximum file descriptors that can be opened by this process.
    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;
     fastcgi_connect_timeout 300;
     fastcgi_send_timeout 300;
     fastcgi_read_timeout 300; 
     fastcgi_buffer_size 64k;
     fastcgi_buffers 4 64k;
     fastcgi_busy_buffers_size 128k;
     fastcgi_temp_file_write_size 128k;
     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;
     #limit_zone  crawler  $binary_remote_addr  10m;
     server
     {
       listen       80;
       server_name  blog.s135.com;
       index index.html index.htm index.php;
       root  /data0/htdocs/blog;
       #limit_conn   crawler  20;   
                          
       location ~ .*\.(php|php5)?$
       {     
         #fastcgi_pass  unix:/tmp/php-cgi.sock;
         fastcgi_pass  127.0.0.1:9000;
         fastcgi_index index.php;
         include fcgi.conf;
       }
  
       location ~ .*\.(gif|jpg|jpeg|png|bmp|swf)$
       {
         expires      30d;
       }
       location ~ .*\.(js|css)?$
       {
         expires      1h;
       }   
       log_format  access  '$remote_addr - $remote_user [$time_local]
 "$request" '
             '$status $body_bytes_sent "$http_referer" '
             '"$http_user_agent" $http_x_forwarded_for';
       access_log  /data1/logs/access.log  access;
       }
       server
      {
       listen       80;
       server_name  www.xxx.com;
       index index.html index.htm index.php;
       root  /data0/htdocs/www;
       location ~ .*\.(php|php5)?$
      {     
       #fastcgi_pass  unix:/tmp/php-cgi.sock;
       fastcgi_pass  127.0.0.1:9000;
       fastcgi_index index.php;
       include fcgi.conf;
      }
      log_format  wwwlogs  '$remote_addr - $remote_user [$time_local]
"$request" '
               '$status $body_bytes_sent "$http_referer" '
              '"$http_user_agent" $http_x_forwarded_for';
   access_log  /data1/logs/wwwlogs.log  wwwlogs;
 }
 server
 {
   listen  80;
   server_name  status.blog.xxx.com;
   location / {
   stub_status on;
   access_log   off;
   }
 }
}

2. 在/usr/local/webserver/nginx/conf/目录中创建fcgi.conf文件:
    vi /usr/local/webserver/nginx/conf/fcgi.conf
    输入以下内容:
    fastcgi_param  GATEWAY_INTERFACE  CGI/1.1;
    fastcgi_param  SERVER_SOFTWARE    nginx;
    fastcgi_param  QUERY_STRING       $query_string;
    fastcgi_param  REQUEST_METHOD     $request_method;
    fastcgi_param  CONTENT_TYPE       $content_type;
    fastcgi_param  CONTENT_LENGTH     $content_length;
    fastcgi_param  SCRIPT_FILENAME    $document_root$fastcgi_script_name;
    fastcgi_param  SCRIPT_NAME        $fastcgi_script_name;
    fastcgi_param  REQUEST_URI        $request_uri;
    fastcgi_param  DOCUMENT_URI       $document_uri;
    fastcgi_param  DOCUMENT_ROOT      $document_root;
    fastcgi_param  SERVER_PROTOCOL    $server_protocol;
    fastcgi_param  REMOTE_ADDR        $remote_addr;
    fastcgi_param  REMOTE_PORT        $remote_port;
    fastcgi_param  SERVER_ADDR        $server_addr;
    fastcgi_param  SERVER_PORT        $server_port;
    fastcgi_param  SERVER_NAME        $server_name;
    # PHP only, required if PHP was built with --enable-force-cgi-redirect fastcgi_param  REDIRECT_STATUS    200;

5. 启动Nginx
    ulimit -SHn 51200
    /usr/local/webserver/nginx/sbin/nginx
配置开机自动启动Nginx + PHP
    vi /etc/rc.local
在末尾增加以下内容:
    ulimit -SHn 51200
    /usr/local/webserver/php/sbin/php-fpm start
    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
使配置立即生效:
    /sbin/sysctl -p
在不停止Nginx服务的情况下平滑变更Nginx配置
1、修改/usr/local/webserver/nginx/conf/nginx.conf配置文件后,请执行以下命令检查配置文件是否正确:
    /usr/local/webserver/nginx/sbin/nginx -t
如果屏幕显示以下两行信息,说明配置文件正确:
    the configuration file /usr/local/webserver/nginx/conf/nginx.conf syntax is ok
    the configuration file /usr/local/webserver/nginx/conf/nginx.conf was tested
    successfully

2、这时,输入以下命令查看Nginx主进程号:
    ps -ef | grep "nginx: master process" | grep -v "grep" | awk -F ' ' '{print $2}'
屏幕显示的即为Nginx主进程号,例如:
    6302
这时,执行以下命令即可使修改过的Nginx配置文件生效:
    kill -HUP 6302
或者无需这么麻烦,找到Nginx的Pid文件:
    kill -HUP `cat /usr/local/webserver/nginx/nginx.pid`
编写每天定时切割Nginx日志的脚本

1、创建脚本/usr/local/nginx/sbin/cut_nginx_log.sh
    vi /usr/local/nginx/sbin/cut_nginx_log.sh
输入以下内容:
    #!/bin/bash
    # This script run at 00:00
    /usr/local/webserver/nginx/sbin/nginx
优化Linux内核参数
    vi /etc/sysctl.conf
在末尾增加以下内容:
    net.ipv4.tcp_fin_timeout = 30
    net.ipv4.tcp_keepalive_time = 300
    #  The Nginx logs path
    logs_path="/usr/local/webserver/nginx/logs/"
    mkdir -p ${logs_path}$(date -d "yesterday" +"%Y")/$(date -d "yesterday" +"%m")/
    mv ${logs_path}access.log ${logs_path}$(date -d "yesterday" +"%Y")/$(date -d "yesterday" +"%m")/access_$(date -d "yesterday" +"%Y%m%d").log
    kill -USR1 `cat /usr/local/webserver/nginx/nginx.pid`

2、设置crontab,每天凌晨00:00切割nginx访问日志
    crontab -e
输入以下内容:
    00 00 * * * /bin/bash  /usr/local/nginx/sbin/cut_nginx_log.sh

0

顶一下

0

踩一下
友情链接 | 诚聘英才 | 关于我们 | 版权声明 | 联系我们 | 广告服务

  •     京ICP备08010289号    Powered by D.V.B