早上到公司查看nginx的错误日志，发现有很多这样的东东：
2012/10/10 19:30:23 [error] 21782#0: *1180689552 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 211.140.18.117, server: xxxx, request: "POST /a.php HTTP/1.1", upstream: "fastcgi://unix:/var/fastcgi-1.sock:", host: "xxxx"
2012/10/10 19:30:23 [error] 21780#0: *1180689995 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 119.39.249.49, server: xxxx, request: "POST /a.php HTTP/1.1", upstream: "fastcgi://unix:/tmp/fastcgi-1.sock:", host: "xxxx"
改域名的访问日志也有很多的502
丫的很久没看见错误日志了先解决呗。查看使用php-cgi的个数，发现没有出现不够用的。仔细看nginx的错误日志这不是php-cgi不够用啊，这是个什么东东呢，网上一通查也没找到答案。唉，还是仔细看看日志吧
发现访问日志中那时候访问a.php的时间都在3s以上。
查看php-cgi.conf原因重要找到了
<value name="request_terminate_timeout">3s</value>
明天继续观察
