[arin-ppml] IPv6 Non-connected networks

[George Bonser]

> 
> Now, the other tier-1's don't like that one bit. Not one bit at all.
> Dude's stealing their bread! So, they start a joint venture to do the
> same thing, camp it at the major peering points (like MAE-East) and
> refuse to honor the /8 route if its announced from anybody else. Sorry
> dude! You want service in the small assignments, you pay the joint
> venture too.
> 

Or, someone starts accepting the more specific routes without using the
tunnels and their customers don't need to pay no steeenking tunnel
broker which disrupts the entire scheme, another network does it, too,
in order to compete with them and that entire tunnel jv blows up and
everyone ends up accepting the more specifics.

What it does is allows those who are willing to make an investment in
routers with more resources (at least at their peering points) and can
handle huge routing tables to be more desirable providers and it drives
competition/innovation among the equipment vendors to ship gear that is
able to handle larger routing tables because the vendors that do so get
the business from the networks that want to offer that kind of service.

This is really going to come to a head with people who want to
dual-stack v4/v6 when v6 routes really start taking off.  Most companies
will be in both spaces. So on one hand we will see more smaller prefixes
in v4 at the same time v6 prefixes are ramping up.  Vendors better be
working on providing gear with a lot more resources for routing and
route caching than they are now.

The major constraint right now on routing table size is really a
hardware issue.  We are all working around a hardware bottleneck that is
just going to get worse.  The vendor to market with routers that handle
substantially larger routing tables without a huge increase in cost is
going to be a winner, and it pays to be a winner.