[arin-ppml] Q1 - ARIN address transfer policy

[Scott Leibrand]

Owen DeLong wrote:
> 
> On Jun 21, 2008, at 4:49 PM, Scott Leibrand wrote:
> 
>> In the case of
>> 2008-2, the goal is to keep deaggregation pressures the same as they
>> would be under a continuation of the current system (if we had more
>> addresses to give out).  In other words, I believe that the
>> extrapolation of current routing table size trend lines that Schiller
>> et. al. have presented represents the eventual situation under 2008-2
>> about as well as possible given the inherent uncertainties.
> 
> I don't believe for one second that 2008-2 as it is currently written would
> achieve this.  I believe it will reduce the speed with which deaggregation
> and disaggregation pressures increase, but, in any market scenario,
> disaggregation pressures will increase.  Some worse than  others
> and APNICs proposal is certainly worse than 2008-2.

Yes, in any exhaustion scenario, deaggregation pressures *on existing 
space* will increase somewhat: as far as I can tell that's inevitable. 
However, I'm not convinced that the rate of increase in the routing 
table will necessarily increase.  Currently we are "deaggregating" the 
free pool into one route per growing multihomed organization per year. 
If we instead deaggregate deployed space to meet the same demand, and 
apply restrictions as in 2008-2, I'm not convinced we'd see any more 
deaggregation than is the inevitable result of more networks joining the 
Internet (and that's a good thing).

Do you have reason to believe that deaggregation pressures will be 
larger under 2008-2 than under a no-transfers exhaustion scenario, where 
incumbents with space lease/SWIP it to downstreams needing the space?

As Randy has pointed out, most of the deaggregates in the current table 
are more-specifics of RIR-allocated routes.  Under a no-transfer or 
2008-2-type policy, I would expect that to remain the case, at least 
unless/until routing table growth starts to outstrip router capacity growth.

-Scott