[106095] in North American Network Operators' Group
Re: virtual aggregation in IETF
daemon@ATHENA.MIT.EDU (Adrian Chadd)
Sun Jul 20 12:34:38 2008
Date: Mon, 21 Jul 2008 00:34:12 +0800
From: Adrian Chadd <adrian@creative.net.au>
To: Joel Jaeggli <joelja@bogus.com>
In-Reply-To: <4883677D.5060509@bogus.com>
Cc: nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
On Sun, Jul 20, 2008, Joel Jaeggli wrote:
> Software switched routers have little pressure on fib limitions. For a
> certain class of application the software switched edge router is in a
> much better position to accommodate fib growth than a device with a
> fixed sized cam.
I dunno about that; there's some papers floating about which look at
trie type FIB representations which note significant savings in compressing
FIB to unique entry set. Less memory, less comparsions, less nodes, etc.
Rather interesting stuff.
Try http://www.academypublisher.com/jnw/vol02/no03/jnw02031827.pdf for
fun.
Adrian
