Re: [tcpm] TCP tuning

[Jerry Chu <[e-mail]>]

On Wed, Feb 3, 2010 at 8:13 AM, Kacheong Poon <[e-mail]> wrote:
> On 02/ 3/10 11:17 PM, SCHARF, Michael wrote:
>
>> In the past, I have played a little bit with an initial window e. g. of
>> 10 MSS, and that worked surprisingly well. Also, one could combine a
>> larger initial window with other mechanisms that would reduce the
>> aggressiveness of the flow startup. Maybe it is indeed time to think
>> about the Slow-Start once again.
>
>
> FWIW, I've come across traces showing some network equipment
> uses an initial window as large as 16 segments (some kind of
> turbo mode).  It is probably true many stacks already use a
> much larger initial window than recommended in the RFC.  Just
> that people have been very quiet about this.

Indeed, the percentage of non-compliance can be quite high, as high
as ~15% in one dataset reported by an IMC paper "TCP Revisited: A
Fresh Look at TCP in the Wild"

Jerry

>
>
>
> --
>
>                                                K. Poon.
>                                                [e-mail]
>
> _______________________________________________
> tcpm mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/tcpm
>