[arin-ppml] Clarify /29 assignment identification requirement

[William Herrin]

On 5/4/12, John Curran <jcurran at arin.net> wrote:
> On May 4, 2012, at 2:33 PM, William Herrin wrote:
>> What I assert is that much of the community would like ARIN to verify
>> utilization using best available approaches at its discretion
>> -excluding- those which would have ARIN consume PII about downstream
>> ISP customers holding fewer than 8 IP addresses.
>
> If that's the case, you should then propose policy which precludes
> ARIN from requesting that information.  It's simple, clear, and can
> be discussed by the community in a straightforward manner.

John,

As you wish.

> Instead, you have proposed a process change that ARIN can request
> this information, but only as last resort. Given that requesting
> this information is the typical method for verifying utilization
> of networks consisting of customer reassignments, I've asked you
> several times what alternatives you propose for verifying the
> utilization of these networks.

At least three such methods have been proposed. One of them - a
statistical analysis on the data - by you.


> Those coming in for address space do not seem to have a problem
> providing customer reassignment information

This discussion started when several folks (none of them me)
complained about providing /32 reassignment information on NANOG. As
you well recall.


> You should propose policy which precludes ARIN from requesting such
> information if that is what you desire. We would then take customer
> redelegations as valid as presented, and that should address your
> concern.

Automatically treating redelegations valid as presented in no way
addresses my concern. In my opinion, any claim that is suspect or
easily faked should be interrogated with the vigor you currently
apply.

It's one particular method you've begun to use which I think steps
beyond the bounds of healthy behavior.


> The only other option that has been clearly defined would be for ARIN
> not ask for individual reassignments to customers but instead to consider
> proof of infrastructure to be sufficient validation for an assertion of
> IP utilization typical for such equipment.  This is appears to be a
> workable policy change to address your concern.

I would think that any indirect evidence of a customer's existence
would support the address utilization claim. A bank statement listing
deposits and charges. A demonstration of access to the routers implied
by assignment claims with apparent programming and interface statuses
that match. A 10q or 10k filing listing customer counts. Really, I
could go on for paragraphs about what sort of anonymous, indirect data
could reasonably imply downstreams' existences in the claimed
quantities. And so can you.

Regards,
Bill Herrin


-- 
William D. Herrin ................ herrin at dirtside.com  bill at herrin.us
3005 Crane Dr. ...................... Web: <http://bill.herrin.us/>
Falls Church, VA 22042-3004