apache core dump

[Clement Laforet]

On Sat, 1 Jan 2005 10:31:55 -0800
"whitevamp" <whitevamp47 at hotmail.com> wrote:

> 
> ok the otherday i had an issue with apache not binding to port 80
> 
> and i thought i got that taken care of, well maby i have but any way
> i knoticed that apache wasnt running any more so i when to start it
> and it whouldnt start so i looked in the error log didnt find
> anything wrong in there so i looked in messages and i found this
> Dec 31 19:00:04 vampextream kernel: pid 88555 (httpd), uid 80: exited on signal 11
> 
> so now its core dumping .. 
> so my quistions are 
> 1. how do i go about debugging this?
First of all you need to find what is causing the segfault:
	- a modules (apache's one or third party one)
	- apache itself

You have to enable or disable every module (one by one) to find the 
culpit. After that, rebuild apache and the culpit module (if any) with 
debugging stuff. After that, you need to allow apache to write a dump.
sysctl kern.sugid_coredump=1
and use CoreDumpDirectory to define a place to store dump files.
run gdb on the core to find where it fails.

> 2. and what would be causeing this?
Generally, mod_php ;)
Find the culpit first, and then you'll see.

clem

Note: if you use nss_ldap + php4, it can cause segfaults.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-apache/attachments/20050103/ec14cd1d/attachment.bin