Sprint's route filters and Europe

Geoff Huston gih at aarnet.edu.au
Tue Jun 18 10:40:10 UTC 1996


errrk - do we all really want to engage in this public resource economics
discussion here yet for yet another round?

Yes we apparently do!

fair enuff.

Public resource economics differs from normal market distribution economics
in that the consideration that charging for a public resource can be an effective
tool in ensuring equitable distribution and equitable exploitation opportunity
as a premium to the cost of the production and distribution infrastructure
associated with the public distribution of the resource.

Of course this is not always the case, and typically a public resource distribution
space couples regulation with a tariff to achieve the ultimate outcome of fair
and equitable distribution. The radio spectrum is perhaps the best covered
territory here when looking at this space in relation to the policy debate over
IP address management.

At the moment we are using the currently unallocated pool of IPV4 address
space as the major policy lever here. as long as this pool is viewed as being
sufficiently large to accomodate future demand modulo the current allocation
practices then this is probably a workable policy application lever (after all that's
the only lever we have left!). However a "once and forever" allocation policy
of release of space from this pool effectively predicts an inevitable demise of this
management mechanism given the nature of the rapacious demand the
market is exerting on this resource of the unallocated space.

My view is that moderation of the registries' policies to whatever extent
we devise is at best a very interim measure which never addresses the nature
of the problem in the long run.

We are left with a scenario where the IPV4 address space will inevitably be a
traded commodity, and the value of the space will be determined by the
percieved value of the address modulo issues of routeability, alternative
mechanisms of dynamic address translation and the perceived value of
deployment of IPv6 as an alternative to securing V4 space.

Thanks,

  Geoff

(trying to keep a quite complex subject as short and pithy as possible)







More information about the NANOG mailing list