[ppml] Policy Proposal: IPv4 Transfer Policy Proposal

Scott Leibrand sleibrand at internap.com
Wed Feb 20 21:50:18 EST 2008


On 02/20/08 at 6:15pm -0800, Leo Vegoda <leo.vegoda at icann.org> wrote:

> On 20/02/2008 17:37, "Owen DeLong" <owen at delong.com> wrote:
>
>>>
>>> I don't know whether 24 months is better than another length of time. 
>>> However, it seems that the current proposal would mean that if a 
>>> network got a six month supply of IPv4 space from ARIN six months 
>>> before the IANA free pool was exhausted they would have to wait a 
>>> further 18 months before they could transfer away any prefixes they no 
>>> longer needed. These numbers would change if proposal 2007-22 is 
>>> accepted.
>>>
>> If they just got a 6 month supply of addresses, shouldn't that imply 
>> that they don't have any addresses they can free up?
>
> There could be addresses to free up when customer numbers drop or service
> offerings change.
>
> In the former case, a drop in revenue could be temporarily plugged by
> 'selling off' address space. In the latter case, an ISP might want to offer
> a basic web-only service provided with an RFC 1918 address while 'selling
> off' the freed up address space to gain some extra revenue.
>
> I think both these scenarios could happen. They aren't necessarily good
> things - but someone else might need that address space and if so, why not
> let them get hold of it?

I would agree, but would also argue that the shorter we make the waiting 
period, the more incentive we give for someone who might be inclined to 
"game" the system, by acquiring as many addresses as they can justify 
prior to exhaustion (causing a "run on the bank") and then turning around 
to transfer them again right after exhaustion.

So, does a 24 month waiting period appropriately balance the legitimate 
need for address recipients to become transferors with the need to prevent 
"flipping"?  Or should that timeframe be shorter/longer?  (I know Leo 
already answered "I don't know", but perhaps others have some input.)

-Scott



More information about the ARIN-PPML mailing list