Re: (ITS#4187) slapd crash

[ando@sys-net.it]

On Sun, 2005-11-20 at 21:17 +0000, kevins@bmrb.co.uk wrote:

> Yes, only simple bind, I haven't even got SASL compiled.  I understand
> your reasons for suggesting an upgrade, but these are my production
> systems and so I'll need to spend some time testing 2.3.12 to make sure
> it all works as expected before I can roll it out.  Unfortunately I
> can't work out how to force the error to occur so the only way I'm going
> to get it tracked down is by monitoring my production systems.
> 
> Would I be okay running 2.3.12 on some boxes and 2.3.11 on others?  I
> could possibly just upgrade my backup production box without having an
> emergency if anything didn't work...

yes.

> 
> >  In case
> > it's not, please try to gather a stack backtrace from a core dump, so
> > that we can try and locate the exact point of failure.
> > 
> I'm happy to do that, either now or on 2.3.12, but could you give me a
> clue where slapd would dump core and whether there is anything special I
> need to do to get that to happen.  I suspect its trying to dump in /,
> but doesn't have write permission - does that sound correct?

I have no clue either.  I suggest you make sure you're running an
unstripped binary (compiled with -g would be great) that can generate a
core (see ulimit(1) with bash), and run it from a terminal keeping it
joined (i.e. run with -d something, even 0).

p.




Ing. Pierangelo Masarati
Responsabile Open Solution

SysNet s.n.c.
Via Dossi, 8 - 27100 Pavia - ITALIA
http://www.sys-net.it
------------------------------------------
Office:   +39.02.23998309          
Mobile:   +39.333.4963172
Email:    pierangelo.masarati@sys-net.it
------------------------------------------