Re: [RE] ClassA traffic pacing approaches
Max,
As the author of the referenced detailed paper, let me
(attempt to) clarify. These reflect my perceptions as
a study group participant; other opinions may vary.
>> In the detailed paper by David James there seems
>> to be a "gating" approach.
The "gating" approach is _a_ solution that can is well
tested on IEEE 1394 and can easily be shown to suffice.
They may be others.
There is consensus that "something" is needed to
shape/pace end-point stations.
Opinions vary on whether shaping/pacing is needed
within bridges.
My recent edits to the bursting/bunching annex scenarios
(might) help to clarify concerns and alternatives.
RE is a study group (not a WG), so downselection
of specific pacing rotocols/parameters is premature.
>> Does this mean that this is the consensus at the
>> moment?
There is no consensus on whether pacing/gating is
appropriate within bridges.
>> In general, is this group trying to establish a
>> specific recommended/mandated pacing approach or
>> just some mandatory performance requirements for
>> this matter?
The goal is to mandate something that can be proven to
work within supported topologies, although the range of
"supported topologies" remains somewhat subjective.
Perhaps this can be done via performance requirements,
perhaps specific algorithms must be specified.
Regardless, we can't rely on "all from one vendor"
installation assumptions.
Cheers,
DVJ
-----Original Message-----
From: owner-stds-802-3-re@ieee.org [mailto:owner-stds-802-3-re@ieee.org]On
Behalf Of Max Azarov
Sent: Monday, June 06, 2005 6:46 AM
To: STDS-802-3-RE@listserv.ieee.org
Subject: [RE] ClassA traffic pacing approaches
Hello all,
Browsing through the available documentation I couldn't get a feeling on
which traffic pacing algorithms are under the consideration in the RE group.
In the detailed paper by David James there seems to be a "gating" approach.
Does this mean that this is the consensus at the moment? In general, is this
group trying to establish a specific recommended/mandated pacing approach or
just some mandatory performance requirements for this matter?
Thanks in advance for clarification,
Max Azarov.