Re: additional requirement: cooperation

[Daniel Senie <dts@senie.com>]

At 01:42 PM 11/12/01, you wrote:
>On 2001-11-12 10:34:25 -0800, Michel Py wrote:
> > These tunnels are annoying in the sense that they could be a loophole if
> > someone wanted to bypass the requirements.
> >
> > What about:
> >
> >     Two sites are "direct neighbours" if they are able to exchange
> > **  traffic betwen their sites directly, without IPv6 datagrams being
> >     processed by a router located at some third site.
> >
> > > "A multihoming strategy MAY require cooperation between a site
> > > and its transit providers, but MUST NOT require cooperation
> > > directly between the transit providers except for direct neighbors".
>
>How about we leave the "except" off and revisit it if it seems really
>necessary later on?  I worry that any protocol that requires
>coordination between competitors is doomed to failure.  :(

I was going to make a similar comment. Further, I will think of Multi6 as a 
failure if the only solution developed does wind up requiring such 
cooperation between competitors. Such requirement fails customers' needs in 
a very important way: failure of one upstream's network should NOT impact 
the viability of connectivity through other upstreams. While this 
cooperative approach is used today, it has generated a significant division 
between the "haves" and the "have nots" separating those who were around 
early and got portable IP space, and those who came later. It is my belief 
that such inequity must not occur in the future.

Should we ultimately decide there is no other choice than upstream 
cooperation, then ALL sites which need to multihome MUST be subject to the 
same rules, and the same pain.

-----------------------------------------------------------------
Daniel Senie                                        dts@senie.com
Amaranth Networks Inc.                    http://www.amaranth.com