RE: [RPRWG] Scheduling
First, I also agree that supporting three classes of service
at the MAC is sufficient.
Multi-word names have been previously used for labelling
classes of service in the context of other standards
efforts, e.g. DiffServ. I don't think the parsing problem is
too hard if we limit ourselves to at most two words, e.g.
"Best Effort". Another option would be to use acronyms.
Although I don't particularly like the concept of using
acronyms, IEEE has previously created ones like "CSMA/CD",
so a few two-letter acronyms that simplify referring to
classes of service in the document would not be too bad.
I am strongly opposed to using the words "synchronous" or
"asynchronous" when labelling traffic classes. The word
"synchronous" would imply some kind of global clock
synchronization to most people, even if it is being used
in a different context here. On the other hand "provisioned"
is a good choice.
Gunes
> Date: Sun, 08 Apr 2001 17:39:44 -0700
> From: "David V. James" <davidvja@xxxxxxxxxxx>
> Subject: RE: [RPRWG] Scheduling
> To: stds-802-17@xxxxxxxx
> Cc: stds-802-17@xxxxxxxx
> MIME-version: 1.0
> Content-transfer-encoding: 7bit
> Importance: Normal
> X-MSMail-Priority: Normal
> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600
> X-Priority: 3 (Normal)
> X-Resent-To: Multiple Recipients <stds-802-17@xxxxxxxxxxxxxxxxxx>
> X-Listname: stds-802-17
> X-Info: [Un]Subscribe requests to majordomo@xxxxxxxxxxxxxxxxxx
> X-Moderator-Address: stds-802-17-approval@xxxxxxxxxxxxxxxxxx
>
>
> All,
>
> I have seen a variety of emails, which share the opinion of the following:
>
>
> -----Original Message-----
> From: owner-stds-802-17@xxxxxxxx [mailto:owner-stds-802-17@xxxxxxxx]On
> Behalf Of Sushil Pandhi
> Sent: Wednesday, April 04, 2001 2:03 PM
> To: William Dai
> Cc: stds-802-17@xxxxxxxx
> Subject: Re: [RPRWG] Scheduling
>
>
> I think 3 priorities seem reasonable: 1) 1 for best effort, 2) 1 for
> guarnateed but non-SYNCH traffic and 3) 1 for SYNCH traffic and for RPR
> Network Control messages.
> -Sushil
> <<<<< End of message <<<<<
>
> I happend to agree with this conclusion.
> But, I would like to strive toward a common
> taxonomy (i.e. glossary) for these three classes.
>
> As an editor, I like to have three distinct one-word names.
> Terms like "best effort" have the problem that they are hard
> to parse when included in sentences. I'll throw out a few
> ideas and ask if anyone has any better ideas:
>
> synchronous - A class of provisioned traffic with guaranteed bandwidth and
> lowest latency.
>
> asynchronous - A clsss of provisioned traffic with guaranteed bandwidth and
> bounded latency.
>
> residual - A class of traffic limited to consumption of unprovisioned or
> unused provisioned bandwidth. Note that RPR attempts to divide residual
> bandwidth fairly among nodes with queued residual traffic.
>
>
> These names are not too bad, but I am a bit concerned that the term
> asynchronous applied to residual as well. Perhaps the terms should be:
>
> Provisioned synchronous
> Provisioned asynchronous
> Unprovisioned asynchronous
>
> But, that seems to be a bit too verbose.
>
> Any other ideas non naming?
>
> DVJ
>
>
>