[RPRWG] Downstream shaper, major deficiencies

["David V James" <dvj@xxxxxxxxxxxx>]

Dongmei,

I'll try to read your concers with the downstream shaper credits
in more detail, but it sounds like this relatively easily managed
by a change in limits or resets when nothing is ready to be sent.

However, I believe there is a far more major problem with the
downstream shaper: it excludes STQ and add of classB. This can
cause violation of classA0 guarantees, due to the cumulative
effect of partially filled upstream STQs after transient loading
conditions.

I have summarized this concern in the attached pdf writeup.
Also listed are other problems related to the coupling of
fairness-eligible and upper-class traffic. I have not yet had
the time to write them in the same level of detail.

I believe the basic premise of our fairness protocols is not
bad, but there are significant remaining problems to be solved.
Unfortunately, its hard to see these through the implementation
details of Clause 9. I hope my writeups will help.

For members of the FAH, this fulfills one of my action items:
>> On the topic of subclassA0.
>> Action items:
>>   DVJ,Stein,Jon to coolaborate on scenario
>>   Stein&Jon to coolaborate on simulation
I hereby hand off this scenario for Stein&Jon review...

DVJ

David V. James
3180 South Ct
Palo Alto, CA 94306
Home: +1.650.494.0926
      +1.650.856.9801
Cell: +1.650.954.6906
Fax:  +1.360.242.5508
Base: dvj@xxxxxxxxxxxx  

>> -----Original Message-----
>> From: owner-stds-802-17@xxxxxxxxxxxxxxxxxx
>> [mailto:owner-stds-802-17@xxxxxxxxxxxxxxxxxx]On Behalf Of Dongmei Wang
>> Sent: Friday, June 20, 2003 6:41 AM
>> To: Jon Schuringa
>> Cc: chwang; stds-802-17@xxxxxxxx
>> Subject: Re: [RPRWG] A problem about shperD in Draft 2.2
>> 
>> 
>> 
>> Jon/Chao:
>> I just noticed this discussion. I have implemented the shaperD 
>> in my simulator,
...

cls09_fairness.pdf