[routing-wg] FW: discussion about rogue database objects

Hello,

----
On Wed, Nov 12, 2014 at 11:18:25AM +0400, poty at iiat.ru wrote:
> I'm not sure the problem is imminent first hand.

The problem is real, and needs to be fixed.
----
Your word against my word - nothing changed in the world.

----
> I'm sure that the routing is not RIPE NCC's business.

As long as nobody else is providing a high-quality and
well-authenticated IRR DB for RIPE region objects, it *very much* is the
RIPE NCC's business.
----
It's not me who mentioned "region". As soon as the authority of RIPE NCC
is limited to RIPE region it is none of the RIPE NCC business to rule
"around the world". I'm in doubt that the burden of inter-RIR arguments
should be put on the end users. RPKI, when it was initially introduced,
was, is and I hope will be independent decision of the end-user.

----
> I believe that the routing object should be created at the place where

> it can be checked properly.

Please provide suggestions how that will help RIPE region LIRs properly
document their routing policy if out-of-region networks are involved.
----
The "routing policy" should be built in each region by their respective
rules. If someone (including RIPE NCC) does not like the rules outside
the region there should be some agreements between RIRs, not between end
users.
I hope you've heard my suggestion to form the routing policy in the
RIR's database where it can be authenticated internally (against ASs and
IP address blocks), than between complicated inter-RIR transfers. In
that way it is only the inter-RIR process of exchanging high-integrity
databases, not multi-thousand companies activity for implementing
untrusted mechanisms to be in place. In my opinion there has been enough
money spent on this already.

Regards,
Vladislav