Re: AppleTalk broadcast addr on FDDI

Rob Peglar <robp@anubis.network.com> Mon, 19 April 1993 23:02 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa27690; 19 Apr 93 19:02 EDT
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa27686; 19 Apr 93 19:02 EDT
Received: from cayman.cayman.com by CNRI.Reston.VA.US id aa12904; 19 Apr 93 19:02 EDT
Received: by cayman.Cayman.COM (4.1/SMI-4.0) id AA22198; Mon, 19 Apr 93 18:13:35 EDT
Return-Path: <robp@anubis.network.com>
Received: from nsco.network.com by cayman.Cayman.COM (4.1/SMI-4.0) id AA22194; Mon, 19 Apr 93 18:13:30 EDT
Received: from anubis-e1.network.com by nsco.network.com (5.61/1.34) id AA04843; Mon, 19 Apr 93 17:15:18 -0500
Received: from landru.network.com by anubis.network.com (4.1/SMI-4.1) id AA14540; Mon, 19 Apr 93 17:12:27 CDT
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Rob Peglar <robp@anubis.network.com>
Message-Id: <9304192212.AA14540@anubis.network.com>
Subject: Re: AppleTalk broadcast addr on FDDI
To: Robert_Jeckell@3mail.3com.com
Date: Mon, 19 Apr 1993 17:12:24 -0500
Cc: raj@doelztc.timeplex.com, apple-ip@cayman.com
In-Reply-To: <930419.140706@3Mail.3Com.COM> from "Robert_Jeckell@3mail.3com.com" at Apr 19, 93 12:16:00 pm
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Content-Length: 5886

Bob et al,

Just a comment from here.

When Bob said:

> Below is the text of the only document I know of that Apple has produced 
> describing FDDITalk.  I'm not sure if a newer version is available.  I haven't 
> heard anything in a long time.

he is right, officially speaking.  However, back in 1990 (March), there
was a group of 4 people at Apple, (including Alan Oppenheimer), 3 at
Cisco, and one from here (NSC) who had an "unofficial FDDITalk specification
meeting".  At this meeting, we decided to use the 802.5, or TokenTalk,
specification as a base for FDDITalk.

This led to the use of the FDDITalk broadcast address of C0-00-04-00-00-00,
which is the same as the TokenTalk broadcast address.  

Our products (and cisco's) have since had to either incorporate a knob to
choose which one to use (NSC) or make people upgrade (cisco) in order to
interoperate with other equipment which uses the 1991 FDDITalk spec.

Everyone now should be using the 1991 spec, which as discussed is 
09-00-07-FF-FF-FF.

Thanks
Rob


> 
> It's funny you should ask at this time.  We implemented to the Apple spec, but 
> I have just recently heard rumors that some third parties (as you have 
> described) don't comply and that there may be some problems out in the real 
> world.  I have asked Apple (Garry Hornbuckle), for any information they have, 
> but I haven't heard back yet.  If I do, I will share the information.
> 
> Unfortunately we didn't go out and buy every FDDI adapter available for our 
> tests.  We used Network Peripheral's adapters on Novell servers when we did our 
> tests last year.  I believe we also did some testing with cisco boxes.  We 
> didn't have any reason to question Apple's document.
> 
> This may be a case of early implementations that cropped up in lieu of a 
> published standard from Apple.  But you would think that implementations would 
> have caught up with the Apple spec.  It's been out for 2 years.  
> 
> By the way, I looked around up on AppleLink this morning and didn't find the 
> FDDITalk document.  I also looked around on both ftp.apple.com.  Anyone at 
> Apple know where it is located these days?
> 
> Has anyone else come across problems?  Any comments?
> 
> /bob
> 
> -----------------------------------------------------------------
> Date: Thu, 27 Feb 1992 08:54:41 -0800
> To: apple-ip@apple.com, mm@hobbit.gandalf.ca
> From: veizades@apple.com
> Subject: Re: fdditalk
> -----------------------------------------------------------------
> The following is the text of the document FDDITalk that is found on
> AppleLink.
> 
> I hope it helps those who need it.
> 
> John...
> 
> --------------------------------------------------------------
> 
> FDDITalk: Preliminary Proposal
> Apple Computer, Inc.
> March 8, 1991
> 
> This document represents a proposed definition for how AppleTalk protocols
> should be implemented on FDDI (that is, for "FDDITalk").  This proposal
> should be considered as highly preliminary and subject to change.  Apple
> will be publishing an official definition of FDDITalk at some point in the
> near future.
> 
> The proposed definition of FDDITalk is simple and easy to state: FDDITalk
> is precisely the same as EtherTalk Phase 2.  Although this is a slight
> oversimplification, it is true from an overall point of view.  To be more
> accurate, FDDITalk packet formats, at the 802.2 level, are precisely the
> same as those used by EtherTalk.  Specifically, AppleTalk data packets and
> AppleTalk Address Resolution Protocol (AARP) packets on FDDI have exactly
> the same format as those on Ethernet (802.3) when looked at from the LLC
> level (MAC headers will of course be different).
> 
> FDDITalk packets, like EtherTalk packets, are encapsulated in SNAP headers.
>  AppleTalk data packets use SNAP type $080007809B.  AARP packets use SNAP
> type $00000080F3.  The AppleTalk broadcast address used by FDDITalk is the
> same as that used by EtherTalk, $090007FFFFFF.  The zone multicast
> addresses are also the same, $090007000000 through $0900070000FC.  Finally,
> the AARP hardware type used by FDDITalk is the same as that used by
> EtherTalk, one (1).
> 
> At this time, it has not been decided whether AARP timeouts on FDDITalk
> will remain the same as those on EtherTalk or not.  The issue of AARP
> timeouts on FDDI is currently under investigation.
> 
> It is important to note that the actual AppleTalk data (i.e. the DDP
> packet) is encapsulated in the FDDITalk packet (that is in the SNAP frame)
> in the same way as it is encapsulated in EtherTalk packets.  This means
> that the maximum size packets used by AppleTalk on FDDI is in the 600 byte
> range (600 bytes plus SNAP, LLC and MAC headers).  The FDDI data link,
> however, provides the ability to transmit multiple back-to-back packets on
> one token, up to the limit placed by the "token hold timer" (a
> MAC-layer-settable parameter).  Through judicious use of this feature, it
> should be possible to achievement end-to-end AppleTalk throughput on FDDI
> roughly equivalent to that which could be achieved through use of much
> larger packets.
> 
> FDDITalk is being defined to be the same as EtherTalk due to the emergence
> of FDDI-to-Ethernet "transparent" bridges.  These bridges make possible
> communication between Ethernet (802.3) and FDDI, at the data link level. 
> Although communication at this level may or may not be desirable (routing
> is often be a better approach), this communication may be necessary in some
> environments. Specifically, environments which already have FDDI-Ethernet
> bridges installed may desire communication between AppleTalk nodes on both
> media without requiring the installation of a router.
> 
> 


-- 
Rob Peglar               Network Systems Corporation
Internetwork Group       7600 Boone Avenue North
robp@anubis.network.com  Minneapolis MN 55428   (612)424-4888 x1028