[802.3_DMLT] Minutes of the IEEE 802.3br (DMLT Task Group) February 13, 2014 teleconference
Minutes of IEEE 802.3br (DMLT Task Group) February 13, 2014 Teleconference
1. Attendees were reminded to comply with the IEEE patent policy.
2. Most of the time has been spent discussing IET negotiation process.
This was done in preparation for March face-to-face meeting. Main
participants in this discussion were Pat Thaler Norm Finn and Hugh
Barass. Norm’s proposal is incorporated into the minutes (see item
4).
3. Pat’s and Hugh’s proposals about pros and cons of using LLDP will be
discussed at the face-to-face meeting.
4. Norm’s opinion on the principles of negotiating the turn-on of IET:
1. The negotiation for one direction is independent of the
negotiation for the other direction. Whether I can transmit to
you is s separate question from whether you can transmit to me.
Any negotiation whose goal is to flip a switch that turns the
feature on or off for both transmission and reception at one end
(e.g, a "turn on the feature in this chip" switch) is
a bad idea. This is because it requires both sides doing
something at the same time, which is inherently difficult. It
inherently requires a firm commitment to action in the future,
which is never easy to achieve with a protocol.
2. A device that has the capability of receiving both non-IET frames
and IET frames is enabled for receiving both kinds of frames at
all times. "Be liberal in what you receive and conservative in
what you send" is a good idea. It eliminates half of the need for
negotiation. Eliminating negotiation is better than doing
negotiation, no matter how clever.
3. The following scenario is a good basis to start with, in my
opinion:
A. As mentioned, if I have IET capability, I always enable
reception of both IET and non-IET frames.
B. After link down, I turn off IET transmission mode.
C. We define an information element for IET negotiation. This
could well be transmitted in 802.1AB LLDP. But, whether
transmitted in LLDP or by hardware negotiation, or by some new
protocol, doesn't matter. What matters is the information.
D. The information transmitted has three parts:
i. I have the ability to receive IET. (The fact that I'm
sending the information is sufficient; we don't need a
0/1 bit.)
ii. I have, or have not (0/1) turned on my ability to
transmit IET.
iii. I have, or have not, recently received an IET frame on
this link.
E. Every time I transmit a PDU, I send the above state
information.
F. If I receive a PDU indicating that you have the ability to
receive IET, I ensure that my IET transmission mode is turned
on. If I ever receive an indication that you do not understand
IET, I turn off IET mode in my transmitter. (I don't need to
do anything with reception mode -- it's always on.)
G. If I have sent you a lot of IET frames, and I have received
several PDUs from you that indicate that you have not received
any IET frames, then something is terribly wrong (e.g. There is
a silent non-IET device between us), and I turn off IET (and
bit ii of the PDU).
Purely for diagnostic purposes, we may add a fourth information
element:
Iv. I have not received anything from you indicating that
you are able to receive IET.
Next teleconference will take place Thursday, February 20, 2014. WebEx
link and dial-in numbers are the same.
Participants:
Ludwig Winkel
Pat Thaler
Norm Finn
Hugh Barass
Albert Tretter
Don Sexton
Alon Regev
Gyorgy Kalman
Helge Zinner
Guest (?)
Anatoly Moldovansky