[84378] in North American Network Operators' Group
Re: OT - Vint Cerf joins Google
daemon@ATHENA.MIT.EDU (Daniel Roesen)
Mon Sep 12 10:11:49 2005
Date: Mon, 12 Sep 2005 16:11:18 +0200
From: Daniel Roesen <dr@cluenet.de>
To: nanog@nanog.org
Mail-Followup-To: nanog@nanog.org
In-Reply-To: <17189.26198.336533.946213@roam.psg.com>
Errors-To: owner-nanog@merit.edu
On Mon, Sep 12, 2005 at 06:28:22PM +0700, Randy Bush wrote:
> those who see full stats at ixes, v4/6 isps, etc will tell you that
> actual v6 traffic is miniscule.
Not contesting the quantification, but what typical IXP switches can
do stats based on ethertype? Given that most relevant IPv6 players
especially in Europe and ASPAC do run dual stack and IPv4 and IPv6
over the same interface, I fail to see how you can see the IPv6 traffic
levels. I can remember Equinix refusing to allow IPv6 on the normal
switch port because they want to see IPv6 traffic levels and need extra
port for that. Which hinders IPv6 deployment even more... extra ports do
cost real extra money for most folks. Perhaps this policy has changed
nowadays (I hope).
Best regards,
Daniel
--
CLUE-RIPE -- Jabber: dr@cluenet.de -- dr@IRCnet -- PGP: 0xA85C8AA0
