Bad routes update

[Alain Durand]

At 15:25 21/07/99 +0400, kuznet@ms2.inr.ac.ru wrote:
>Hello!
>
> > I think thrashing out these routing policy issues and actually trying them
> > in practice is one of the best uses of the 6bone. It's certainly not much
> > good for getting packets from A to B :-).
>
>Yup, it is truth 8)8)
>
>Thing, which I do not understand is how you want to investigate
>routing policy with <256 policed entities involved? It cannot result
>in any realistic results. That's my main point. You have to explode
>6bone to make it closer to reality, rather than shrink it.
>
>Alexey

We have been operating the 6 bone for about 2 years
with BGP4+ without enforcing any policy. This gave us enough
time to understand what can/will go wrong with route announcement.
All this experience lead us to RFC2546 and the current hardening draft.

It is now time to move from an experimental network to a pre-production
one. This means some kind of quality and robustness.
This is in my opinion an important step forward to get ISP on board
and convince them to offer native v6 service to their customers.

         - Alain.