[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::fddi

Title:FDDI - The Next Generation
Moderator:NETCAD::STEFANI
Created:Thu Apr 27 1989
Last Modified:Thu Jun 05 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2259
Total number of notes:8590

203.0. "FDDI Addressing" by BERN01::DEY (Walter Dey, EIS, Berne Switzerland) Wed Feb 13 1991 03:38

Stupid question: What is the MAC address of a host directly connected to a WC ?
for TCP/IP, DECnet Ph IV and V. I know that there is a RFC for FDDI, but
didn't have time to study it it. I have no idea where to find any info about
DECnet ?

Walter.
T.RTitleUserPersonal
Name
DateLines
203.1KONING::KONINGLietuva laisva!Wed Feb 13 1991 16:3054
That's not a stupid question at all.  There are several parts.

1. Whether or how you're connected to a concentrator is irrelevant.    
   A concentrator is a physical layer device.  The Ethernet analogy
   would be "For a node connected to a DEMPR, what's the address...".

2. You need to distinguish the station address (known in FDDI terminology
   as "MLA" or "My Long Address") vs. "alias" addresses.
   In Ethernet we also make that distinction, but only a few of the
   adapters support it  (QNA/LQA, BNI/MNA, SGEC).  In FDDI, ALL adapters
   are required to support it.

The MLA is taken from the address ROM.  You cannot change it.
The MLA is the DEFAULT address for each of the protocols that
use the adapter.

For each individual protocol, when you open the socket, or channel, on
the adapter, you have the option to specify what individual address you
want to use.  If you don't supply one, the MLA is used.  If you do specify
one, that one is used instead; we call that an "alias address".  Note that
each protocol can independently supply an alias address (a different one,
if you want).  The alias address does NOT cause the MLA to change.

To be specific: most protocols do not ask for an alias address.  The major
one that does is DECnet Phase IV, which asks for AA-00-04-00-xx-yy where
yyxx is the 16 bit node address.

Once this is done, the data link (adapter plus driver) will look for
packets of the appropriate protocol type, with the specified individual
address.  For example, if my MLA is 08-00-2B-12-00-14, and my DECnet Phase IV
address is 1.2 (0402 hex), and I also turn on TCP/IP, the data link would
be looking for, among others:

	DA = 08-00-2B-12-00-14, protocol = 08-00 (IP)
	DA = AA-00-04-00-02-04, protocol = 60-03 (DECnet IV)

Exactly the same answer applies to Ethernet, with one additional issue.
Some Ethernet adapters are not able to look for more than one individual
address at the same time.  In that case, the only choice we have is to
use as the station address the address required by the FIRST data link
user that is enabled.  If others then make conflicting requests, and
error results.  This is why you currently have to turn on DECnet first
and LAT next.

The reason we came up with the concept of the multiple alias addresses is to
eliminate that problem with order of startup.  The reason why we REQUIRE
the support in FDDI, and don't allow MLA to be changed under any 
circumstance, is that some of the basic functions of FDDI (such as the
claim token process, which creates the token) depend on MLA, and in
particular on having MLA be unique.  By not allowing MLA to change, we
ensure that network management mistakes cannot break the claim token
process.

	paul