Re: Again no multi6 at IETF#56

[David Conrad <david.conrad@nominum.com>]

Hmmm.

If you believe this to be an appropriate function of the RIRs, feel 
free to propose it for an RIR public policy meeting.

Bring marshmallows... :-)

Rgds,
-drc

On Monday, March 17, 2003, at 02:11  PM, Masataka Ohta wrote:
> drc;
>
> > Define "tier 1".
> Good question.
>
> Let RIRs define it.
>
> It can, for example, be defined purely technically with # of POPS and
> so on or purely financially with auction.
>
> 						Masataka Ohta
> > Rgds,
> > -drc
> >
> > On Monday, March 17, 2003, at 12:11  PM, Masataka Ohta wrote:
> >
> > > Michel;
> > >
> > > > > Joe Abley wrote:
> > > > > To be pedantic, the requirements doc doesn't specify
> > > > > the starting point
> > > > I do not agree. It is clear from that paragraph that the starting
> > > > point
> > > > is either "Current IPV4 multihoming practices" or "the Internet's 
> > > > BGP
> > > > Routing Table".
> > > Not at all.
> > >
> > > The global routing table should contain one entry for each tier1
> > > ISP or, maybe, a little more than that.
> > >
> > > Initially, there may be 10.
> > >
> > > AS numbers, of course, may be assigned to tier2 and other ISPs,
> > > which have multiple peers to other ISPs and want to control
> > > traffic with BGP.
> > >
> > > 							Masataka Ohta