[26138] in North American Network Operators' Group
Re: Verio Decides what parts of the internet to drop
daemon@ATHENA.MIT.EDU (Alex P. Rudnev)
Mon Dec 6 02:15:42 1999
Date: Mon, 6 Dec 1999 10:04:12 +0300 (MSK)
From: "Alex P. Rudnev" <alex@virgin.relcom.eu.net>
To: Yakov Rekhter <yakov@cisco.com>
Cc: nanog@merit.edu
In-Reply-To: <199912051353.FAA28924@omega.cisco.com>
Message-ID: <Pine.SUN.4.10.9912060959360.13359-100000@virgin.relcom.eu.net>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Errors-To: owner-nanog-outgoing@merit.edu
I mean that, if there is rule no allow some prefixes, all ISP should follow this
rule. All! It's the issue.
Else, we have just what we have tiday - everything work, but some stohastic
filtering add a lot of unpredicted behaviour into the routing. I know a lot of
such specifics (class B / 24, see 144.206.* for example) which (1) work OK, (2)
surely some part of the routing follow 144.206/16, but at some point /24
speciics appear and routing became just as it should be).
Today, the only 2 rules are:
(1) some blocks (195.* etc) can't use more than /20 (npot /19) specifics; it
began a few years ago due to the Sprint policy;
(2) Less than /24 announces are not mainly allowed.
If someone want more (and no doubt many ISP want), let's they follow some
negotiated rulesa and ask another to follow this rules, too.
Alex R (now not ISP network engeneer, through).
On Sun, 5 Dec 1999, Yakov Rekhter wrote:
> Date: Sun, 05 Dec 1999 05:53:57 -0800
> From: Yakov Rekhter <yakov@cisco.com>
> To: Alex P. Rudnev <alex@virgin.relcom.eu.net>
> Cc: nanog@merit.edu
> Subject: Re: Verio Decides what parts of the internet to drop
>
> Alex,
>
> > may be - but it shoudl be written in the RFC, not in the VERIO's policy. The
> > global policy must be THE SAME over the global Internet.
>
> What would make such a policy "THE SAME over the global Internet" ?
>
> Yakov.
>
Aleksei Roudnev,
(+1 415) 585-3489 /San Francisco CA/
