What to do with Adjacency table until we make the fix going to draft

saperia@tcpjon.ogo.dec.com Sat, 17 October 1992 23:04 UTC

Message-Id: <9210172221.AA22219@tcpjon.ogo.dec.com>
To: rndi!rnd!rina@uunet.uu.net
Cc: phiv-mib@pa.dec.com, saperia@tcpjon.ogo.dec.com
Subject: What to do with Adjacency table until we make the fix going to draft
Date: Sat, 17 Oct 1992 18:21:04 -0400
Sender: ietf-archive-request@IETF.NRI.Reston.VA.US
From: saperia@tcpjon.ogo.dec.com

Segdan,
A couple of things:

	1.  I have added you to the phiv-mib distribution list so that you can
talk with others who are interested and are working on similar implementations.

	2.  I have forwarded your note to the group since I wanted to have
them have an opportunity to disagree whith my answer to you about what to do
in the interim.

	3.  I think it is in everyone's best interest if we agree on an
interim implementation approach until the RFC is fixed.

	4.  My response:

You observe correctly that the RFC as currently written has a problem with the
Adjacency Index.  This mailing list had a fairly log discussion about what to
do.  I will forward you separately, the final proposal for a fixed table.  In
short, we plan to obsolete the current table and put in a new table which is
nearly identical to the current table except that its index  is both the
circuit index and the node addr.  After much discussion, we concluded that
this was the best fix even though there were less radical approaches such as
simply having a unique index for each adjacency.  While we could do this as an
interim approach, I am not sure that this is the best thing to do.  This is
why I am posting this here.  My gut reaction is to suggest that people
implementing the MIB use the new table that we proposed which will be placed
at the end of the Adjacency group.  I think we will get best interoperability
this way and have minimal impact when the document is revised.  It also will
ensure that we do not have OID problems.

Comments?????

/jon
------- Forwarded Message

Return-Path: rndi!rnd!rina@uunet.UU.NET
Received: by tcpjon.ogo.dec.com (5.57/ULTRIX-fma-071891);
	id AA10388; Thu, 15 Oct 92 21:03:20 -0400
Received: by inet-gw-2.pa.dec.com; id AA15046; Thu, 15 Oct 92 18:00:20 -0700
Received: from uunet.uu.net (via LOCALHOST.UU.NET) by relay2.UU.NET with SMTP (5.61/UUNET-internet-primary) id AA23246; Thu, 15 Oct 92 20:59:07 -0400
Received: from rndi.UUCP by uunet.uu.net with UUCP/RMAIL(queueing-rmail) id 205837.2119; Thu, 15 Oct 1992 20:58:37 EDT
Received: by rndi (5.65/1.35)id AA00415; Fri, 16 Oct 92 01:26:01 +0200
Received: by  (4.1/SMI-4.1)id AA02914; Thu, 15 Oct 92 12:33:41 IST
Date: Thu, 15 Oct 92 12:33:41 IST
From: rndi!rnd!rina@uunet.UU.NET (Rina Nethaniel)
Message-Id: <9210151033.AA02914@>
To: saperia
Subject: DecNet MIB


Dear sir,

        We are currently implementing the DECnet Phase IV MIB in our management
station. while doing so we came across a problem with the adjacency table. The a
djacency table uses circuit index as a key but the adjacency entry is defined as
 "one entry in the table for each adjacency". In case of Ethernet circuits the r
esult is a group of adjacencies that can vary from 1 to 1023. While given an Eth
ernet circuit no. what will be returned from the router ?

        Thank you
                yours sincerely
                Segdan Refaela
                Rad Network Devices
                ATIDIM Technological Park
                Bldg. # 1  Tel-Aviv , Israel



------- End of Forwarded Message

What to do with Adjacency table until we make the… saperia