Wierd applicatioin hang

I have an application that I piggybacks off of our RT’s fastcgi/mason
infrastructure.
I have recently ported it to run against RT 4.0, on rhel, using an IBM
release of apache 2.4, and now, I can only run one request on my
application code before the the server hangs. I do not encounter the hang
problem if I only use RT, without accessing my app.

The hang results in a 500 error to the browser, and the apache error log
entries below.

Has anybody run into this before? Any ideas on how to troubleshoot.

For what it is worth, when I shut down apache, the fast cgi process(es) are
never shut down as well.

Len.

[Tue Aug 20 12:03:55 2013] [info] [client 10.13.11.215] [ed002470] Session
ID: d1cAAI8HZD9PWXOrc7R9FotqsE5YWFhYMpMTUgAAAAA= (reused)
[Tue Aug 20 12:04:09 2013] [notice] mpmstats: rdy 44 bsy 6 rd 5 wr 1 ka 0
log 0 dns 0 cls 0
[Tue Aug 20 12:05:09 2013] [notice] mpmstats: rdy 44 bsy 6 rd 5 wr 1 ka 0
log 0 dns 0 cls 0
[Tue Aug 20 12:06:09 2013] [notice] mpmstats: rdy 44 bsy 6 rd 5 wr 1 ka 0
log 0 dns 0 cls 0
[Tue Aug 20 12:07:09 2013] [notice] mpmstats: rdy 44 bsy 6 rd 5 wr 1 ka 0
log 0 dns 0 cls 0
[Tue Aug 20 12:08:10 2013] [notice] mpmstats: rdy 44 bsy 6 rd 5 wr 1 ka 0
log 0 dns 0 cls 0
[Tue Aug 20 12:08:55 2013] [error] [client 10.13.11.215] FastCGI: comm with
server “/u01/app/PROD/rt4-efts/sbin/rt-server.fcgi” aborted: idle timeout
(300 sec)
[Tue Aug 20 12:08:55 2013] [error] [client 10.13.11.215] FastCGI:
incomplete headers (0 bytes) received from server
“/u01/app/PROD/rt4-efts/sbin/rt-server.fcgi”
[Tue Aug 20 12:08:55 2013] [info] [client 10.13.11.215] [ed100468] Session
ID: dVcAALtZlZDW6kcjnv5D5CVPwQtYWFhYl5QTUgcAAAA= (new)

lenjaffe@jaffesystems.com 614-404-4214 @lenjaffe
Software Engineer, Founder Volunteerable http://www.volunteerable..com/,
Host of Columbus Code Jam http://www.meetup.com/techlifecolumbus/
www.lenjaffe.com http://www.theycomewithcheese.com
Perl Advent Planet http://www.lenjaffe.com/PerlAdventPlanet/ - Advent
Calendars: Perlish and otherwise.