[3427] in linux-net channel archive
Re: BW limiting/simulation
daemon@ATHENA.MIT.EDU (Alan Cox)
Sun Jun 23 15:27:25 1996
From: alan@lxorguk.ukuu.org.uk (Alan Cox)
To: dennis@etinc.com (dennis)
Date: Sun, 23 Jun 1996 17:46:15 +0100 (BST)
Cc: alan@lxorguk.ukuu.org.uk, linux-net@vger.rutgers.edu
In-Reply-To: <199606231617.MAA14792@etinc.com> from "dennis" at Jun 23, 96 12:17:16 pm
[Discussion about precise techniques removed - I think you'll find Dennis we
were arguing the same thing]
> You're assuming that every TCP implementation in the world works the way you
> think
> it does. Believe me this is a very dangerous assumption. You also have to
> account for
> other types of traffic (other than TCP), so this alone is not adaquate, even
> if it works.
I'm assuming every TCP implementation in the world backs off on losing frames.
If it doesn't its not a TCP implementation. You also don't depend on the
behaviour of the other TCP beyond that because if it adjusts more slowly it will
take several drops before it gets the speeds right.
The simulation with any traffic shaper is very close to precisely the behaviour
you would get from an actual physical device. The problem is purely one of
efficient representation of the simulation rather than "what are we simulating"
Its also a topic with enough literature to give you a years reading.
Alan
