[122436] in North American Network Operators' Group
Re: History of 4.2.2.2. What's the story?
daemon@ATHENA.MIT.EDU (Joe Abley)
Sun Feb 14 18:04:42 2010
From: Joe Abley <jabley@hopcount.ca>
To: Mark Andrews <marka@isc.org>
In-Reply-To: <201002142243.o1EMhx44071536@drugs.dv.isc.org>
Date: Sun, 14 Feb 2010 18:04:46 -0500
X-SA-Exim-Mail-From: jabley@hopcount.ca
Cc: Sean Reifschneider <jafo@tummy.com>, "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
On 2010-02-14, at 17:43, Mark Andrews <marka@isc.org> wrote:
> Using three consecutive addresses doesn't remove
> single points of failure in the routing system.
That depends on how the routes for those destinations are chosen, and
what routing system you're talking about.
For distribution of a service using anycast inside a single AS, and
with one route per service, it makes no difference whether the
addresses are adjacent. Two /24 routes are no more stable than two /32
routes within an IGP. There's no prefix filtering convention to
accommodate, here.
>
>> If their goal is distribute a service for the benefit of their own =
>> customers, then keeping all anycast nodes associated with that
>> service =
>> on-net seems entirely sensible.
>
> Which only helps if *all* customers of those servers are also on net.
Whether it helps depends on what Level3's goals are. This is not
public infrastructure; this is a service operated by a commercial
company.
For what it's worth, I have never heard of an ISP, big or small,
deciding to place resolvers used by their customers in someone else's
network. Perhaps I just need to get out more.
Joe
