[Iccrg] Re: draft-fairhurst-tcpm-newcwv-03 - Status?

Michael Welzl michawe at ifi.uio.no
Fri Jun 22 08:28:27 BST 2012


Hi,

First of all, the draft is here:
http://tools.ietf.org/html/draft-fairhurst-tcpm-newcwv-03

I think there were no public comments on the mailing list, just at the 
meeting @ IETF 80. Here they are:
http://www.ietf.org/proceedings/80/minutes/iccrg.txt

As for the question "may we now return to TCPM": we don't take a formal 
role in this process - i.e., you may decide to take this to TCPM at any 
time.

The point of ICCRG review is to get comments and incorporate feedback, 
up to a point where, if someone in TCPM says "this and that looks like 
open research, you should take this to ICCRG" you could ideally point to 
reviews that were already done there, and comments that were 
incorporated, thereby hopefully settling the discussion. If pointing to 
reviews can't settle it, this is going to make the document bounce back 
to ICCRG with a specific request from TCPM, which is also fine, but I 
think it would be most efficient to minimize (or, even better, 
altogether avoid) such iterations.

Since Gorry is saying that they'd love more comments, can I ask for 
volunteers to do a thorough review? For previous proposals of changes to 
TCP's congestion control, we had 2-3 such reviews.

Cheers,
Michael


On 6/21/12 4:43 PM, Gorry Fairhurst wrote:
> ICCRG,
>
> We have revised our draft called "Updating TCP to support
> Variable-Rate Traffic". We think the feedback we received after ICCRG
> has been addressed and are grateful to all who commented. This new
> version is also intended to be easier to read.
>
> Here is a summary of issues and NiTs that were corrected:
> (1)    Non-validation phase time changed to 5 from 6 minutes,
>     following comments that TCP aleady has a 5 minute timer.
> (2)    Section 4.3.2:  At the end of the recovery phase, the TCP sender
>     MUST reset the cwnd to (FlightSize-R)/2 and not FlightSize/2.
> (3)     Various places to improve language and address questions about
>     ambiguities after iccrg.
> (4)     Note change of author (Israfil has now moved-on to a job
>     elsewhere).
>
> We'd love more comments, and also like to know what happens next to
> this draft. Are there more things that ICCRG wishes to think upon, or
> may we now return to TCPM with a proposal to take this forward there?
>
> Best wishes,
>
> Gorry
>
>
>
>
>
>
>





More information about the Iccrg mailing list