[GRLUG] fastcgi problem

[Thad Ward]

While I haven't used nginx, and only played with php via fastcgi a little,
it looks like php might be dying for some reason (syntax error, timeout,
request limit reached). After getting a 502, check to see if php is still
running.

Some solutions I have found involve running php under spawn-fcgi using
daemontools:

http://wiki.linuxwall.info/doku.php/en:ressources:dossiers:nginx:daemontools_spawnfcgi

Another solution is to monitor the php process yourself and restart it, but
it is not as elegant as letting daemontools do the monitoring for you:

http://hostingfu.com/article/keeping-your-php-fastcgi-processes-alive

Someone else suggested using TCP rather than unix sockets, but I'm not
convinced that's why it is dying.

--- On Fri, 7/22/11, L. V. Lammert <lvl at omnitec.net> wrote:

From: L. V. Lammert <lvl at omnitec.net>
Subject: [GRLUG] fastcgi problem
To: "Grand Rapids LUG" <grlug at grlug.org>
Date: Friday, July 22, 2011, 12:54 PM

Have a roundcube server that has developed an issue today:

2011/07/22 11:50:57 [error] 24722#0: *1061075 connect() to
unix:/tmp/php.sock failed (61: Connection refused) while connecting to
upstream, client: 206.197.251.252, server: mail.middiv.com, request: "GET
/index.php HTTP/1.1", upstream: "fastcgi://unix:/tmp/php.sock:", host:
"mail.middiv.com"


It appears to be breaking the connection to fastcgi, .. and nginx reports
a 502 Bad Gateway.

Have not used fastcgi enough - any suggestions on how to isolate the
issue?


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.