问题
Things like this keep popping up in my apache logs. How can I go about debugging this and/or preventing this in the future:
(104)Connection reset by peer: FastCGI: comm with server "/var/www/php5.external" aborted: read failed
FastCGI: incomplete headers (0 bytes) received from server "/var/www/php5.external",
(4)Interrupted system call: FastCGI: comm with server "/var/www/php5.external" aborted: select() failed
回答1:
In relation to your "select() failed" error and possibly the others, check your version of libapache2-mod-fastcgi There were some issues that have since been patched.
Upgrading libapache2-mod-fastcgi from 2.4.6 to 2.4.7 seems to rectify this issue.
More info here:
http://geektank.net/2008/11/fastcgi-and-dreaded-aborted-select-failed/
http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/3de22bc415d3da02?pli=1
However, I'm not sure what OS/Apache version you are using, but hopefully this helps you solve these issues.
回答2:
I was getting errors 1. and 2. today on libapache2-mod-fastcgi 2.4.7~0910052141-1.
In my case, I found something that helped - oddly enough not in the correct virtualhost server logfile, but in the general one (/var/log/apache2/error.log):
[Thu Jun 05 16:41:47 2014] [error] server is within MinSpareThreads of MaxClients, consider raising the MaxClients setting
I'm still debugging why exactly that happened, this answer here being of great help.
来源：https://stackoverflow.com/questions/9586094/apache-php5-fpm-connection-reset-by-peer