[28791] in North American Network Operators' Group
RE: Private ASN suppression
daemon@ATHENA.MIT.EDU (Dmitri Krioukov)
Tue May 16 16:13:11 2000
From: "Dmitri Krioukov" <dima@dimension.net>
To: <danny@tcb.net>, "Howard C. Berkowitz" <hcb@clark.net>
Cc: <nanog@merit.edu>
Date: Tue, 16 May 2000 16:25:42 -0400
Message-ID: <NCBBIKACLKNMKDHKKKNFCEGDEKAA.dima@dimension.net>
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
In-Reply-To: <200005161844.MAA28881@tcb.net>
Errors-To: owner-nanog-outgoing@merit.edu
> -----Original Message-----
> From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu]On Behalf Of
> Danny McPherson
> Sent: Tuesday, May 16, 2000 2:45 PM
> To: nanog@merit.edu
> Subject: Re: Private ASN suppression
>
>
>
>
> In short, see RFC 2270.
...as well as rfc2260.
to be more specific, we have to note that two different options
are considered there. if you use the first one (read rfc),
then you cannot use private asn until you're ok with
generating inconsistencies (and it seems from the previous
discussions of this topic that this becomes (illegal? ->
no answer...) practice for some smaller isps).
with the second option of 2260, you can use private asn
since the more specific pa routes are always aggregated
(and you cannot use confederations, btw).
after all, as was noted, remove-private-as on cisco can be
replaced by a simple route map, and the attribute manipulation
functionality should definitely exist on versalars...
--
dima.
