[112742] in North American Network Operators' Group
Re: Pro-actively publishing IRR data
daemon@ATHENA.MIT.EDU (Joe Provo)
Mon Mar 16 13:20:16 2009
Date: Mon, 16 Mar 2009 13:20:04 -0400
From: Joe Provo <nanog-post@rsuc.gweep.net>
To: Steve Bertrand <steve@ibctech.ca>
In-Reply-To: <49BE7AD0.30205@ibctech.ca>
Cc: nanog list <nanog@nanog.org>
Reply-To: nanog-post@rsuc.gweep.net
Errors-To: nanog-bounces@nanog.org
On Mon, Mar 16, 2009 at 12:14:08PM -0400, Steve Bertrand wrote:
[snip]
> Are there any potential dangers of publishing our information before we
> use it that I may be overlooking?
In case you are worried about folks who filter, recall that the IRR
uniqueeness is based upon the Prefix/length, originAS, and sourceDB
tuple. There is nothing wrong with registering the origin as it is
parallel to the origin-as-it-will be. If *you* do it, instead of
your provider, you will not have to co-ordinate with anyone to yank
the old one. Purposeful overlapping for the purposes of transition
is one of the features.
Cheers,
Joe
--
RSUC / GweepNet / Spunk / FnB / Usenix / SAGE
