httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Cliff Woolley <cliffwool...@yahoo.com>
Subject Re: apache crash
Date Wed, 11 Jul 2001 01:21:59 GMT
On Tue, 10 Jul 2001, Jack Gostl wrote:

> We am running apache 1.3.17 with mod_ssl under AIX 4.2.1. We have a fast
> CGI app, written in C, that has been running well for app 2 months. Now,
> twice in the last two days, one of the apache tasks suddenly started using
> huge amounts of memory. Eventually all of the paging space is used the
> machine comes crashing down. It wasn't even a particularly busy day.
> Even though the memory allocated was "charged" to apache, I suspect that
> fastcgi was somehow involved.

The first thing I'd do is try to figure out what request it was that
triggered the loop.  [Presumably some particular (possibly bogus) request
tripped it off, as opposed to it having just occurred spontaneously after
working for two months.] Check your access logs and try to figure out what
the commonality was between the two occurrences.  Then attempt to manually
duplicate the problem, but be ready with a debugger and top on-hand, ready
and running.  When the program starts to run away, top should make it
obvious which pid is causing the problem.  Attach to that pid with the
debugger, interrupt the process, and get yourself a backtrace.  That
should help in determining where the problem is happening.

--Cliff

--------------------------------------------------------------
   Cliff Woolley
   cliffwoolley@yahoo.com
   Charlottesville, VA



Mime
View raw message