[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: openldap on alpha (fwd)

This should probably go here at this point.

Dan                              <dan@mid-hudson.com>
Network Systems Engineer          518.943.6600x3350

---------- Forwarded message ----------
Date: Fri, 9 Jun 2000 23:56:35 -0400 (EDT)
From: dan@mid-hudson.com
To: "Kurt D. Zeilenga" <Kurt@OpenLDAP.org>
Cc: dan@mid-hudson.com, openldap-software@OpenLDAP.org
Subject: Re: openldap on alpha

It seems as if a patch has been available since Apr 28th


but it still hasn't been merged into either a debian package or upstream. 
I know that Ben C (bcollins@debian.org) has said on both the debian list


and in the ITS on this issue


that he knows about the issue, but it seem like small patch that could be
quickly merged.  It does *run* without the patch, but the performance hit
on the Alpha DS10 is bad enough that I wouldn't want to put it into
production like that.  I'm currently running with the patch applied by

As it seems to be a small path, although you may want to do it
differently, and its is listed as an important issue that has been open
for almost 45 days, I'm curious as to why it hasn't been added.  I'd be
much more comfortable putting this setup into production if I didn't have
to apply that patch by hand.  I am, by far, a programmer.

Please correct me if I'm off base on this.  The more I learn, the less I

Network Systems Engineer   

On Fri, 9 Jun 2000, Kurt D. Zeilenga wrote:

> Date: Fri, 09 Jun 2000 19:27:36 -0700
> From: "Kurt D. Zeilenga" <Kurt@OpenLDAP.org>
> To: dan@mid-hudson.com
> Cc: openldap-software@OpenLDAP.org
> Subject: Re: openldap on alpha
> At 01:14 PM 6/8/00 -0400, dan@mid-hudson.com wrote:
> >I've installed openldap 1.2.10 on my alpha ds-10s and it seems to run, but
> >every time I query it, it logs a few messages:
> >
> >slapd(5341): unaligned trap at 0000000120038d6c: 000003ffffbffc04 2d 2
> >slapd(5341): unaligned trap at 0000000120038d6c: 000003ffffbffc0c 2d 2
> This is a known bug in OpenLDAP, IIRC, in the BER code.  It *should*
> run correct, albiet slowed by the necessary realignment.
> >Everything seems to function, but I plan on using the database for a
> >number of things, so a few lines of errors on each query is going to be a
> >mess.  Any suggestions or pointers?
> Well, you have a simple choice.  You either live with it or
> fix it.  Terry has some useful suggestions on debugging alignment
> problems.  See ITS/bugs for pass comments regarding this.
> Patches welcomed.  Coordinate/discussion development on the
> devel list.
> Kurt