[67162] in North American Network Operators' Group
Re: Strange public traceroutes return private RFC1918 addresses
daemon@ATHENA.MIT.EDU (Stephen J. Wilcox)
Tue Feb 3 15:55:15 2004
Date: Tue, 3 Feb 2004 20:52:51 +0000 (GMT)
From: "Stephen J. Wilcox" <steve@telecomplete.co.uk>
To: Petri Helenius <pete@he.iki.fi>
Cc: Niels Bakker <niels=nanog@bakker.net>, <nanog@merit.edu>
In-Reply-To: <401FFC3A.8000909@he.iki.fi>
Errors-To: owner-nanog-outgoing@merit.edu
On Tue, 3 Feb 2004, Petri Helenius wrote:
> Stephen J. Wilcox wrote:
>=20
> >>Why large MTU then? Most modern ethernet controllers don=B4t care if yo=
u=B4re
> >>sending 1500 or 9000 byte packets. (with proper drivers taking advantag=
e of
> >>the features there) If you=B4re paying for 40 byte packets anyway, ther=
e is no
> >>incentive to ever go beyond 1500 byte MTU.
> >> =20
> >>
> >
> >I think its partially due to removal of overhead and improvements you ge=
t out of=20
> >TCP (bearing in mind it uses windowing and slow start)
> >
> Sure, if you control both endpoints. If you don=B4t and receivers have sm=
all
> (4k,8k or 16k) window sizes, your performance will suffer.
>=20
> Maybe we should define if we=B4re talking about record breaking attempts =
or real
> operationally useful things here.
By definition of this discussion about using large MTU we are assuming that=
=20
packets are arriving >1500 bytes and therefore that we do have control of t=
he=20
endpoints and they are set to use jumbos
Steve
