www-apache-bugdb mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Lew Payne <le...@netcom.com>
Subject Re[2]: Changed information for PR general/389
Date Thu, 24 Apr 1997 15:42:46 GMT

Recently, you (Randy Terbush) wrote:
>> Synopsis: httpd - caught SIGSEGV (segment violation!)
> Did you get my email earlier this morning?  I would suggest making 
> the change provided by Marc Slemko so that we can get a core dump, 
> and to perhaps grab a copy of the latest dev sources. Many fixes 
> related to signal handling have been applied. It would probably be 
> most informative if you can work with the latest sources. If you 
> cannot easily grab a copy of the latest CVS snapshot from the 
> Apache site, let me know and I will send you a copy.

Hello Randy,
  Yes, I did get your message after I had replied to another.  I also
received (this morning) another complaint from the PHP mailing list.
Basically, he's getting the same thing in a LINUX environment, and it
started after he upgraded to Apache 1.2b8 and CVS PHP.  Unfortunately,
he changed both variables simultaneously, so we don't know which one
did it.  Do you still think this is FreeBSD-related?
  Another message on the PHP group talked about SIGSEGV after issuing
the ClearStatCache() function.  Rasmus has corrected it in the newest
CVS release.  Unfortunately, he also announced his CVS is down for a
few days.  I'm not issuing a ClearStatCache(), so this may be academic.
  Allow me to explain my situation... The machine in question is in a
production environment.  It handles approximately 300,000 visits (and
22 GB) of our 550,000 visitors/day.  In order to upgrade the machine,
I had to hack the NetBSD "de" driver to get FreeBSD to work with my
10-Base-100 DEC chipset ethernet card.  Then I had to graft ifconfig
as well as headers.  It was a royal pain.  I'm switching to the Intel
100B next time I have to drive down to CIX/PAIX again.  I am hesitant
to grab a CVS FreeBSD and play with it on this machine.  Rebuilding
the kernel and rebooting is okay.  Reinstalling everything from scratch
and incurring that downtime isn't.
  Randy, is there an alternative you can think of?  I have yet to
learn how to install a new version of FreeBSD on a running machine
without wiping the disk.  The easiest would be for me to set up
another machine to do this on, but that will take a week or two.

View raw message