【 概述 】
在PHP开发中工作里非常多使用到超时处理到超时的场合,我说几个场景:
1. 异步获取数据如果某个后端数据源获取不成功则跳过,不影响整个页面展现
2. 为了保证Web服务器不会因为当个页面处理性能差而导致无法访问其他页面,则会对某些页面操作设置
3. 对于某些上传或者不确定处理时间的场合,则需要对整个流程中所有超时设置为无限,否则任何一个环节设置不当,都会导致莫名执行中断
4. 多个后端模块(MySQL、Memcached、HTTP接口),为了防止单个接口性能太差,导致整个前面获取数据太缓慢,影响页面打开速度,引起雪崩
5. 。。。很多需要超时的场合
这些地方都需要考虑超时的设定,但是PHP中的超时都是分门别类,各个处理方式和策略都不同,为了系统的描述,我总结了PHP中常用的超时处理的总结。
【Web服务器超时处理】
[ Apache ]
一般在性能很高的情况下,缺省所有超时配置都是30秒,但是在上传文件,或者网络速度很慢的情况下,那么可能触发超时操作。
目前 apache fastcgi php-fpm 模式 下有三个超时设置:
fastcgi 超时设置:
修改 httpd.conf 的fastcgi连接配置,类似如下:
<ol class="dp-xml"> <li class="alt"><span><span class="tag"><span class="tag-name">IfModule</span><span> mod_fastcgi.c</span><span class="tag">></span><span> </span></span></span></li> <li><span> FastCgiExternalServer /home/forum/apache/apache_php/cgi-bin/php-cgi -socket /home/forum/php5/etc/php-fpm.sock </span></li> <li class="alt"><span> </span></li> <li><span> ScriptAlias /fcgi-bin/ "/home/forum/apache/apache_php/cgi-bin/" </span></li> <li class="alt"><span> </span></li> <li><span> AddHandler php-fastcgi .php </span></li> <li class="alt"><span> </span></li> <li><span> Action php-fastcgi /fcgi-bin/php-cgi </span></li> <li class="alt"><span> </span></li> <li><span> AddType application/x-httpd-php .php </span></li> <li class="alt"><span> </span></li> <li> <span class="tag"></span><span class="tag-name">IfModule</span><span class="tag">></span><span> </span> </li> </ol>
缺省配置是 30s,如果需要定制自己的配置,需要修改配置,比如修改为100秒:(修改后重启 apache):
<ol class="dp-xml"> <li class="alt"><span><span class="tag"><span class="tag-name">IfModule</span><span> mod_fastcgi.c</span><span class="tag">></span><span> </span></span></span></li> <li><span> </span></li> <li class="alt"><span> FastCgiExternalServer /home/forum/apache/apache_php/cgi-bin/php-cgi -socket /home/forum/php5/etc/php-fpm.sock -idle-timeout 100 </span></li> <li><span> </span></li> <li class="alt"><span> ScriptAlias /fcgi-bin/ "/home/forum/apache/apache_php/cgi-bin/" </span></li> <li><span> </span></li> <li class="alt"><span> AddHandler php-fastcgi .php </span></li> <li><span> </span></li> <li class="alt"><span> Action php-fastcgi /fcgi-bin/php-cgi </span></li> <li><span> </span></li> <li class="alt"><span> AddType application/x-httpd-php .php </span></li> <li><span> </span></li> <li class="alt"> <span class="tag"></span><span class="tag-name">IfModule</span><span class="tag">></span><span> </span> </li> </ol>
如果超时会返回500错误,断开跟后端php服务的连接,同时记录一条apache错误日志:
<ol class="dp-xml"> <li class="alt"><span><span>[Thu Jan 27 18:30:15 2011] [error] [client 10.81.41.110] FastCGI: comm with server "/home/forum/apache/apache_php/cgi-bin/php-cgi" aborted: idle timeout (30 sec) </span></span></li> <li><span> </span></li> <li class="alt"><span>[Thu Jan 27 18:30:15 2011] [error] [client 10.81.41.110] FastCGI: incomplete headers (0 bytes) received from server "/home/forum/apache/apache_php/cgi-bin/php-cgi" </span></li> </ol>
其他 fastcgi 配置参数说明:
<ol class="dp-xml"> <li class="alt"><span><span>IdleTimeout 发呆时限 </span></span></li> <li><span>ProcessLifeTime 一个进程的最长生命周期,过期之后无条件kill </span></li> <li class="alt"><span> </span></li> <li><span>MaxProcessCount 最大进程个数 </span></li> <li class="alt"><span> </span></li> <li><span>DefaultMinClassProcessCount 每个程序启动的最小进程个数 </span></li> <li class="alt"><span> </span></li> <li><span>DefaultMaxClassProcessCount 每个程序启动的最大进程个数 </span></li> <li class="alt"><span> </span></li> <li><span>IPCConnectTimeout 程序响应超时时间 </span></li> <li class="alt"><span> </span></li> <li><span>IPCCommTimeout 与程序通讯的最长时间,上面的错误有可能就是这个值设置过小造成的 </span></li> <li class="alt"><span> </span></li> <li><span>MaxRequestsPerProcess 每个进程最多完成处理个数,达成后自杀 </span></li> </ol>
[ Lighttpd ]
配置:lighttpd.conf
Lighttpd配置中,关于超时的参数有如下几个(篇幅考虑,只写读超时,写超时参数同理):
主要涉及选项:
<ol class="dp-xml"> <li class="alt"><span><span class="attribute">server.max-keep-alive-idle</span><span> = </span><span class="attribute-value">5</span><span> </span></span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-read-idle</span><span> = </span><span class="attribute-value">60</span><span> </span> </li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.read-timeout</span><span> = </span><span class="attribute-value">0</span><span> </span> </li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-connection-idle</span><span> = </span><span class="attribute-value">360</span><span> </span> </li> </ol>
<ol class="dp-xml"> <li class="alt"><span><span>-------------------------------------------------- </span></span></li> <li><span> </span></li> <li class="alt"><span># 每次keep-alive 的最大请求数, 默认值是16 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-keep-alive-requests</span><span> = </span><span class="attribute-value">100</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># keep-alive的最长等待时间, 单位是秒,默认值是5 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-keep-alive-idle</span><span> = </span><span class="attribute-value">1200</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># lighttpd的work子进程数,默认值是0,单进程运行 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-worker</span><span> = </span><span class="attribute-value">2</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># 限制用户在发送请求的过程中,最大的中间停顿时间(单位是秒), </span></li> <li><span> </span></li> <li class="alt"><span># 如果用户在发送请求的过程中(没发完请求),中间停顿的时间太长,lighttpd会主动断开连接 </span></li> <li><span> </span></li> <li class="alt"><span># 默认值是60(秒) </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-read-idle</span><span> = </span><span class="attribute-value">1200</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># 限制用户在接收应答的过程中,最大的中间停顿时间(单位是秒), </span></li> <li><span> </span></li> <li class="alt"><span># 如果用户在接收应答的过程中(没接完),中间停顿的时间太长,lighttpd会主动断开连接 </span></li> <li><span> </span></li> <li class="alt"><span># 默认值是360(秒) </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-write-idle</span><span> = </span><span class="attribute-value">12000</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># 读客户端请求的超时限制,单位是秒, 配为0表示不作限制 </span></li> <li><span> </span></li> <li class="alt"><span># 设置小于max-read-idle时,read-timeout生效 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.read-timeout</span><span> = </span><span class="attribute-value">0</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># 写应答页面给客户端的超时限制,单位是秒,配为0表示不作限制 </span></li> <li><span> </span></li> <li class="alt"><span># 设置小于max-write-idle时,write-timeout生效 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.write-timeout</span><span> = </span><span class="attribute-value">0</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span># 请求的处理时间上限,如果用了mod_proxy_core,那就是和后端的交互时间限制, 单位是秒 </span></li> <li><span> </span></li> <li class="alt"> <span class="attribute">server.max-connection-idle</span><span> = </span><span class="attribute-value">1200</span><span> </span> </li> <li><span> </span></li> <li class="alt"><span>-------------------------------------------------- </span></li> </ol>
说明:
对于一个keep-alive连接上的连续请求,发送第一个请求内容的最大间隔由参数max-read-idle决定,从第二个请求起,发送请求内容的最大间隔由参数max-keep-alive-idle决定。请求间的间隔超时也由max-keep-alive-idle决定。发送请求内容的总时间超时由参数read-timeout决定。Lighttpd与后端交互数据的超时由max-connection-idle决定。
延伸阅读:
http://www.snooda.com/read/244
[ Nginx ]
配置:nginx.conf
<ol class="dp-xml"> <li class="alt"><span><span>http { </span></span></li> <li><span> </span></li> <li class="alt"><span> #Fastcgi: (针对后端的fastcgi 生效, fastcgi 不属于proxy模式) </span></li> <li><span> </span></li> <li class="alt"><span> fastcgi_connect_timeout 5; #连接超时 </span></li> <li><span> </span></li> <li class="alt"><span> fastcgi_send_timeout 10; #写超时 </span></li> <li><span> </span></li> <li class="alt"><span> fastcgi_read_timeout 10; #读取超时 </span></li> <li><span> </span></li> <li class="alt"><span> </span></li> <li><span> </span></li> <li class="alt"><span> #Proxy: (针对proxy/upstreams的生效) </span></li> <li><span> </span></li> <li class="alt"><span> proxy_connect_timeout 15s; #连接超时 </span></li> <li><span> </span></li> <li class="alt"><span> proxy_read_timeout 24s; #读超时 </span></li> <li><span> </span></li> <li class="alt"><span> proxy_send_timeout 10s; #写超时 </span></li> <li><span> </span></li> <li class="alt"><span>} </span></li> </ol>
说明:
Nginx 的超时设置倒是非常清晰容易理解,上面超时针对不同工作模式,但是因为超时带来的问题是非常多的。
延伸阅读:
http://hi.baidu.com/pibuchou/blog/item/a1e330dd71fb8a5995ee3753.html
http://hi.baidu.com/pibuchou/blog/item/7cbccff0a3b77dc60b46e024.html
http://hi.baidu.com/pibuchou/blog/item/10a549818f7e4c9df703a626.html
http://www.apoyl.com/?p=466
1