Cell Relay Archive

Cell Relay Retreat>List Archive>month:1999-Jun> msg00121



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: ILMI address registration question

  • From: albert.e.manfredi@boeing.com
  • Date: Mon, 21 Jun 1999 17:57:52 GMT
  • Organization: Boeing North American
  • X-Article-Creation-Date: Mon Jun 21 17:57:52 1999 GMT


In article <376E61E2.9B94742F@lucent.com>,
  "ronald h. davis" <ronaldd@lucent.com> wrote:

> what is the g/l bit?  i mean, i think that i get the idea that it is
> to indicate global/local administration but where is it defined?

When transmitting a 48-bit MAC address, the first bit across the wire
is called I/G bit, and the second bit is the U/L bit (I had called it
G/L, but the 802.3 Spec calls it U/L).

The first bit defines the address as being "individual" or "group,"
i.e. a unicast vs a multicast Layer 2 address. 0 = individual address.

The second bit says whether it is locally or globally administered. 0 =
globally administered.

Note that in canonical order, i.e. as these go across an Ethernet
interface, these two bits are in the LSB positions of the first byte
(i.e. to the right). In token ring and FDDI order, they are on the left
of that first byte.

When the updated RFC 1483 shows encapsulations of 802.3 frames, it will
specify canonical order. When it shows 802.5 frames, it will specify
the 802.5/FDDI ordering.

This is all very confusing, but it seems to work. How the MAC would be
specified for the ESI bytes in an ATM address (AESA) I don't know. In
theory, it should be in that 802.5 order, but I much prefer canonical.
It's confusing as hell to decode the OUI (most significant 24 bits of
the MAC address) when it's in reverse order.

--
Bert
manfredi@arl.bna.boeing.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.