[RPRWG] Comments on the draft and Attendance at the upcoming meeting
John Lemon made two very important points in the attached email. I am
sending it to the entire reflector so that everyone reading the present
draft is fully aware of them.
(1) If you have found what you believe to be any deficiencies in the
present draft it is critical that you submit comments in those areas, rather
than assume someone else will. Posting your concern to the reflector in no
way guarantees that it will show up as a comment in the data base. If it
does not show up, then it will not be dealt with.
(2) John's second point is equally important. Although we would greatly
appreciate your attendance and participation at the meeting, lack of
attendance will in no way change the seriousness of addressing your comment.
What it will change is to eliminate the ability of you to clarify what you
meant in your comment. Therefore, please be as complete and as clear as you
can be in pointing out what you believe to be deficiencies in the draft, and
the way you propose to resolve those deficiencies.
Thank you.
Best regards,
Robert D. Love
President, LAN Connect Consultants
7105 Leveret Circle Raleigh, NC 27615
Phone: 919 848-6773 Mobile: 919 810-7816
email: rdlove@xxxxxxxx Fax: 208 978-1187
----- Original Message -----
From: "John Lemon" <JLemon@xxxxxxxxxxxx>
To: "Yan Robichaud" <yan@xxxxxxxxxxxxxxx>
Cc: "David V James" <dvj@xxxxxxxxxxxx>; "Dongmei Wang"
<mei@xxxxxxxxxxxxxxxx>; "Harry Peng" <hpeng@xxxxxxxxxxxxxxxxxx>; "Jon
Schuringa" <jon.schuringa@xxxxxxxxxxxx>; "K. K. Ramakrishnan"
<kkrama@xxxxxxxxxxxxxxxx>; "Khaled Amer" <amer@xxxxxxxxxxx>; "Kshitij Kumar"
<kkumar@xxxxxxxxxxxxxx>; "Leon Bruckman" <leonb@xxxxxxxxxxxxx>; "Necdet
Uzun" <nuzun@xxxxxxxxx>; "Paritosh Kulkarni" <paritosh@xxxxxxxxxxxxxxx>;
"Robert D Love" <rdlove@xxxxxxxxx>; "Stein Gjessing" <steing@xxxxxxxxx>
Sent: Tuesday, July 15, 2003 1:54 PM
Subject: RE: Questions regarding D2.3
Yan,
Unfortunately, while I keep intending to get around to studying these more
closely, I've been too swamped to do more than a cursory glance. So, I would
not be the appropriate person. Unless someone else who has a clear
understanding of the issues volunteers, it would be best for you to submit
the comments so that the best explanation can be given. Your lack of
attendance will not in any prevent the comments from getting their due
attention. (All of us on this list will make sure of that.)
jl
-----Original Message-----
From: Yan Robichaud [mailto:yan@xxxxxxxxxxxxxxx]
Sent: Tuesday, July 15, 2003 10:31 AM
To: John Lemon
Cc: David V James; Dongmei Wang; Harry Peng; Jon Schuringa; K. K.
Ramakrishnan; Khaled Amer; Kshitij Kumar; Leon Bruckman; Necdet Uzun;
Paritosh Kulkarni; Robert D Love; Stein Gjessing
Subject: Re: Questions regarding D2.3
Hi John,
Since I am not going to the july meeting, would it be easier for you or
another person to submit the comments to be able to discuss them in person?
Yan
----- Original Message -----
From: "John Lemon" <JLemon@xxxxxxxxxxxx>
To: "Yan Robichaud" <yan@xxxxxxxxxxxxxxx>; "David V James"
<dvj@xxxxxxxxxxxx>; "Dongmei Wang" <mei@xxxxxxxxxxxxxxxx>; "Harry Peng"
<hpeng@xxxxxxxxxxxxxxxxxx>; "Jon Schuringa" <jon.schuringa@xxxxxxxxxxxx>;
"K. K. Ramakrishnan" <kkrama@xxxxxxxxxxxxxxxx>; "Khaled Amer"
<amer@xxxxxxxxxxx>; "Kshitij Kumar" <kkumar@xxxxxxxxxxxxxx>; "Leon Bruckman"
<leonb@xxxxxxxxxxxxx>; "Necdet Uzun" <nuzun@xxxxxxxxx>; "Paritosh Kulkarni"
<paritosh@xxxxxxxxxxxxxxx>; "Robert D Love" <rdlove@xxxxxxxxx>; "Stein
Gjessing" <steing@xxxxxxxxx>
Sent: Tuesday, July 15, 2003 1:17 PM
Subject: RE: Questions regarding D2.3
> Yan,
>
> I hope you are submitting comments based on these questions against draft
2.3. I suspect we may need to alter some behavior, or at least clarify some
text to correct or answer these. Without comments, we won't have any ability
to do so.
>
> jl
>
> -----Original Message-----
> From: Yan Robichaud [mailto:yan@xxxxxxxxxxxxxxx]
> Sent: Monday, July 07, 2003 3:06 PM
> To: David V James; Dongmei Wang; Harry Peng; John Lemon; Jon Schuringa;
> K. K. Ramakrishnan; Khaled Amer; Kshitij Kumar; Leon Bruckman; Necdet
> Uzun; Paritosh Kulkarni; Robert D Love; Stein Gjessing
> Subject: Questions regarding D2.3
>
>
> Hello everyone,
>
> I have a few questions and comments regarding Draft 2.3, in particular to
> those implementing a simulator.
>
>
> //// Q1 ////
> In Table 9.7 rows 5 and 6, the checks
>
> addRate+fwdRate > highRateThreshold
> addRate+fwdRate < lowRateThreshold
>
> are done in the 1 transit buffer case.
>
> Since the algorithm need to check all unreserved traffic types, why the
> table is not using
>
> nrXmitRate > highRateThreshold
> and nrXmitRate < lowRateThreshold
>
> to also include ClassB-CIR traffic?
>
>
> //// Q2 ////
> I noticed in T9.5-2 p231 that the counters decay is now (compared with
D2.2)
> done before the checks in T9.7. Thus, for example, addRate and fwdRate
are
> now 75% of the value they were in D2.2 when executing the comparison. This
> increase the amount of addRate+fwdRate that can be transited before
becoming
> greater than highRateThreshold.
>
> Is this really the expected behavior?
>
>
> //// Q3 ////
> On p223 l39, forwardedFeFrame is defined as "received" frames, and not
> "forwarded" frames. Also, ComputeActiveWeights on p243 is using the
> function transitedFeFrame. Maybe it would be better to call both functions
> receivedFeFrame.
>
> Now that the local station is always included in the activeWeights
> calculation (p243 l22), in the 1 transit buffer case when a unique greedy
> upstream station sends traffic, the downstream station (which does not add
> any traffic) declares congestion and advertises a fairRate = 0.5 *
> unreservedRate and then slowly increase until it becomes congested and
> advertise 0.5 * unreservedRate again, causing oscillations (assuming all
> weights = 1, thus activeWeights=2).
>
> Is this the expected behavior?
>
>
> //// Q4 ////
> In T9.4-10, isAddRateOk uses nrXmitRate < unreservedRate. However, on p214
> l45 there is no mention of it.
>
> Which is which?
>
>
>
> //// C1 ////
> p222: stqFullThreshold definition missing.
>
> //// C2 ////
> In clause 6, CreditFE is not used anywhere.
>
>
> Thanks in advance,
>
> Yan Robichaud
> carleton University
>
>