[75730] in North American Network Operators' Group
Re: Stupid Ipv6 question...
daemon@ATHENA.MIT.EDU (Kevin Oberman)
Mon Nov 22 12:28:25 2004
To: Joe Abley <jabley@isc.org>
Cc: Lars Erik Gullerud <lerik@nolink.net>,
crist.clark@globalstar.com, Stephen Sprunk <stephen@sprunk.org>,
North American Noise and Off-topic Gripes <nanog@merit.edu>
In-reply-to: Your message of "Sun, 21 Nov 2004 19:55:10 EST."
<295304CA-3C21-11D9-9D98-000D93B24C7A@isc.org>
Date: Mon, 22 Nov 2004 09:27:57 -0800
From: "Kevin Oberman" <oberman@es.net>
Errors-To: owner-nanog-outgoing@merit.edu
> From: Joe Abley <jabley@isc.org>
> Date: Sun, 21 Nov 2004 19:55:10 -0500
>
>
> On 20 Nov 2004, at 19:13, Kevin Oberman wrote:
>
> > In any case, if the prefix length is >64, routing is done in the
> > CPU.
>
> Engineers at Juniper seem to be telling me that this is definitively
> not the case for their M- and T-series routers. Which routers were you
> referring to?
Odd. Juniper engineers have assured me that this is th case with M and T
series routers (or any router using the IP2 chip).
To clarify a bit, if the networks are connected, or "direct" in
Juniper-ese. then the CPU is not involved. Only if there is a real
"routing decision" made. OS if you have several connected /126s or /127s
on a single router, you are OK, but if you are truly sub-netting a
prefix longer than a /64 to several routers, then the CPU gets to figure
out where a packet goes.
I'd love to hear this is wrong, but it was confirmed to m by a rather
senor engineer at Juniper, not a JTAC phone droid. Would Tony care to
comment?
--
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman@es.net Phone: +1 510 486-8634
