[arin-ppml] REQUEST FOR ARIN STAFF Was: Re: Policy Proposal 120: Protecting Number Resources

[Owen DeLong]

On Nov 9, 2010, at 5:41 PM, Leo Bicknell wrote:

> In a message written on Tue, Nov 09, 2010 at 04:35:03PM -0800, Owen DeLong wrote:
>> How does reclamation in any way improve this beyond what is
>> accomplished by marking records with invalid/unverifiable POC
>> data as such and leaving them?
> 
> I believe there are folks who will look for that in whois and target
> those blocks for hijacking.  I'd like to think every ISP carefully
> checked the records before routing things, but we have years of
> evidence that is not the case.  I fear we are creating an attractive
> nussance.
> 
And they won't look for blocks that simply aren't in whois?

I don't buy the argument. The hijackers have no trouble finding the
blocks. This fallacious argument was used for a long time to block
efforts to get these blocks marked in public view. We finally came
to realize that marking them publicly allows for filtration to prevent
hijacking.

Obscuring the existence of the blocks does nothing to stop
hijacking. Marking them visibly does.

>> The only reason to bother reclaiming is if you intend to reissue.
>> Since there does seem to be consensus that reissuing is of little
>> value, is there any reason to allocate resources to reclamation
>> instead of merely marking invalid/unresponsive records?
> 
> I believe the space should be re-issued.  I also disagree strongly
> that reissueing is of little value.  It may be of little value with
> respect to the problem of extending the life of IPv4, but that is
> not the only measure.
> 
While I agree there is minimal value in reissuing, I think it is
dwarfed by the costs of dubious reclamations once the prior
organization notices.

> It will be of huge value to the lucky recipients who are able to
> get the space from ARIN who otherwise would not have had that
> opportunity and either had to go without or buy it on the open
> market.  Several others have attempted to estimate the value using
> various measures, and the numbers are quite large.  I think there
> is likely a lot of value to getting this space back in the system.
> 
Until they end up in a routing war with the prior holder or in
a protracted legal battle they didn't expect to be involved in
as a codefendant to ARIN.

I think that Mr. Hannigan's $40/IP is a rather inflated view of
the value of IPv4. At that price, IPv6 is cost effective almost
regardless of the amount of hardware you have to replace.

Many of the other more spectacular numbers have been even
more inflated (I've heard as much as $1,000 per IP). I simply
don't think they are real.

Owen