[arin-ppml] The Library Book Approach to IPv4 Scarcity

[Stephen Sprunk]

Chris Grundemann wrote:
> == Potential Proposal:
>
> Once every 12months each holder of IPv4 addresses is required to fully 
> document their IP utilization and demonstrate that the current 
> utilization standard for IPv4 assignments and allocations is being 
> met. This shall include all currently held IPv4 space, regardless of 
> origin or registration status.

Other than the period (see below), this is the degenerate case of 
2007-14 -- and fear of that degenerate case is/was a major part of the 
opposition to that proposal.  Like many others, I object to _all_ 
registrants being subjected to this process regularly, regardless of the 
period; it should be up to ARIN staff to decide which registrants 
require this sort of attention -- and where ARIN's (read: our) money is 
most efficiently spent.  Reviewing all that documentation ain't cheap.

> A fee shall be assessed for underutilization or insufficient 
> documentation.

ARIN is not -- and IMHO should not be -- in the business of levying 
fines.  The only real enforcement powers ARIN has are (a) refusing to 
allocate/assign new resources, and (b) revoking existing resources.  I 
am very leery of changing that.

>     * The fee for one 12m period shall be waived if the address holder 
> returns a contiguous block of IPv4 space equal to at least 1/256th of 
> currently held space and no less than one /24 (class C equivalent) to 
> ARINs free pool.
>     * The fee for one 12m period shall be waived if the address holder 
> signs an ARIN RSA for any uncontested and unregistered IPv4 space, 
> this waiver shall be restricted to one use per member organization.

There are existing policies that have a similar purpose which don't 
appear to have had any significant effect to date.  Plus, fees (and any 
waivers thereof) are generally the purview of the BoT.  However, I would 
support the general idea of either of these proposals if they were not 
tied to the above "library fee" proposal.

> 6) I originally considered a period of 24 months but shortened it to 
> 12 months considering the rapid approach of IANA free pool exhaustion; 
> 24 months will be far to long of an interval to have a significant 
> impact on IPv4 availability.

2007-14 originally had a (minimum) 12-month window between reviews, but 
there was a very strong consensus that that was too short, so we 
modified it to 24 months.  I would suggest the same to you so that you 
don't run into the same opposition.  It could be reduced later in a 
separate proposal if it becomes absolutely necessary.

S