[6bone] Re: RFC 2772 input from RIR space holder

[Gert Doering]

Hi,

On Fri, Nov 22, 2002 at 12:40:33AM +0000, Tim Chown wrote:
> On Thu, Nov 21, 2002 at 06:14:22PM +0000, Robert Kiessling wrote:
> > That's an interesting proposal.
> 
> You're assuming the 2001: space is "clean", which with its recent requirement
> relaxation and explosion is allocations, is less likely to be true (as 
> people copy 6bone practice to the "production" space).  

While Robert is assuming this, it does not really matter.

If the 2001:: people mess up their stuff, peer pressure among ISPs
has a good chance of cleaning that up (by dropping peerings and so on),
but it's an "internal" thing - if 6bone messes up 2001:: routing, 
the established mechanisms for communication among ISPs just doesn't
work anymore.

So what Robert's proposal actually does is "encapsulate routing
instabilities inside RIR space / inside 6bone space (wherever it 
originates)".  

If 6bone is unstable, RIR space can still be "productive", and are not
directly affected.

If RIR space is unstable, the RIR space holders have to fix it (and
depending on the way filters will be set up, this might not necesarily
affect 6bone space).

Gert Doering
        -- NetMaster
-- 
Total number of prefixes smaller than registry allocations:  50279 (49875)

SpaceNet AG                 Mail: netmaster@Space.Net
Joseph-Dollinger-Bogen 14   Tel : +49-89-32356-0
80807 Muenchen              Fax : +49-89-32356-299