Re: Using X.500 to determine presentationAddresses

Steve Kille <S.Kille@isode.com> Fri, 14 May 1993 18:28 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa08601; 14 May 93 14:28 EDT
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa08597; 14 May 93 14:28 EDT
Received: from haig.cs.ucl.ac.uk by CNRI.Reston.VA.US id aa18125; 14 May 93 14:28 EDT
Received: from glengoyne.isode.com by haig.cs.ucl.ac.uk with Internet SMTP id <g.04286-0@haig.cs.ucl.ac.uk>; Fri, 14 May 1993 18:53:14 +0100
Received: from glengoyne.isode.com by glengoyne.isode.com with SMTP (PP) id <09403-0@glengoyne.isode.com>; Fri, 14 May 1993 18:55:16 +0100
To: Peter Yee <yee@atlas.arc.nasa.gov>
cc: osi-ds@cs.ucl.ac.uk
Subject: Re: Using X.500 to determine presentationAddresses
Phone: +44-71-721-7582
In-reply-to: Your message of Wed, 05 May 1993 23:57:37 -0700.
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 14 May 1993 18:55:13 +0100
Message-ID: <9401.737402113@isode.com>
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Steve Kille <S.Kille@isode.com>

Peter,

First, I think that this is a very important function.   I think that
a key benefit of X.500 is its flexibility to provide a mix of services
using a uniform naming framework.

The decisions on naming machines should be primarily dictated by
 1) Reasonable names for the machines
 2) A naming structure which permits effective allocation.
I would expect that some key services would be named at the ARC level,
and the majority of machines at the departmental or project level
(i.e., org units within ARC).   

If search performance is a problem, I suggest that the aim should be
to work on the DSA engine.  We should be building tools to provide the
services needed.


Steve Kille

 >From:  Peter Yee <yee@atlas.arc.nasa.gov>
 >To:    osi-ds@cs.ucl.ac.uk
 >Subject: Using X.500 to determine presentationAddresses
 >Date:  Wed, 05 May 1993 23:57:37 -0700

 >I was involved, the other day, in a lively discussion on how we could use our
 >X.500 directory systems to look up presentation addresses for various services
 >and hosts.  I'm not talking about a small number of machines, but potentially
 >thousands.  The questions boiled down to: "How do you name a host and service
 >in the OSI world (akin to a DNS hostname and a TCP/UDP service name/port)?",
 >and given that you have such a name, "what is the best way to use X.500 to
 >retrieve the corresponding presentation address?"  Has anyone given thought
 >to how this should be done?
 >
 >I'm aware that the ISODE contains an implementation of DASE which allows one
 >to look up a presentation address given a user-friendly name (UFN) and an
 >application context.  However, if, for example, I wanted to put all of the
 >machines at NASA Ames Research Center into the directory and use DASE to do
 >lookups, I would end up with a few thousand additional entries under the DIT
 >location US, NASA, ARC.  This is where my people entries are.  And this is
 >where one might logically expect to find my machine entries.  Clumping them
 >together, though, means that search performance for both people lookups and
 >service lookups goes down.  And if I'm not using DASE, I've no idea where I
 >might need to search (particularly since there is no standardized naming
 >scheme for OSI end-systems).
 >
 >Well, has anyone got some ideas on this one.  I'd love to hear them.  Anything.
 >
 >My thanks in advance.
 >
 >							-Peter Yee
 >							yee@atlas.arc.nasa.gov