[arin-ppml] Draft Policy ARIN-2013-6: Allocation of IPv4 and IPv6 Address Space to Out-of-region Requestors - Revised Problem Statement and Policy Text

[Matthew Petach]

On Mon, Sep 16, 2013 at 3:59 PM, David Farmer <farmer at umn.edu> wrote:

> On 9/14/13 22:58 , Matthew Petach wrote:
>
>  Why not simply use a phrase like "significant fraction" rather than
>> "plurality"?
>>
>
> The problem with significant fraction, its overly vague, while plurality
> may not be a commonly use every day word, it does have a precise meaning
> and in this context that is "more than any other RIR's region". However,
> since there are 5 regions the smallest possible plurality would be slightly
> more than 20% within the ARIN region.  However, in most cases a plurality
> will be more than that.
>
> Rather than significant fraction, if the community could agree on a
> percentage say 20%, 25%, or maybe 30%, as a minimum percentage within the
> region that would be a little simpler than plurality, and be actually
> something staff could implement. I do not believe significant fraction as
> the standard would give staff a policy that can be implemented.
>
>

Can you clarify if this policy only applies
to *new* requests, or if it is meant to apply
to *all* number resources requested from
ARIN?

I have a serious problem with trying to assign
a fixed value if this is meant to apply to existing
as well as new number resources.  But if it it only
applies to new requests, and the requirement for
a plurality of the *new* request being used to service
needs somewhere within the ARIN region, I see that
as less onerous than the notion of a plurality of *all*
number resources having to be within the ARIN
region.



>
>  change
>>
>> " a plurality of resources
>> requested from ARIN must be justified by technical infrastructure and
>> customers located within the ARIN service region, and any located
>> outside the region must be interconnected to the ARIN service region."
>>
>> to
>>
>> " a significant fraction of the resources requested from ARIN must be
>> justified by technical infrastructure or customers located within the ARIN
>> service region, and any located outside the region must be interconnected
>> to the ARIN service region."
>>
>
> If we don't like plurality for whatever reason, I'd suggest;
>
> "a minimum of X% of the resources requested from ARIN must be justified by
> technical infrastructure or customers located within the ARIN service
> region, and any located outside the region must be interconnected to the
> ARIN service region."
>
> Where X% is something like 20%, 25%, or 30%.


So, how about something like this, then?

"a minimum of 20% of the *new* resources requested from ARIN must be
justified by technical infrastructure or customers physically located within
the borders of ARIN member countries, and any technical infrastructure
or customers located outside the ARIN region must be physically
interconnected
to the ARIN service region"



>
>
>  (representing a global network that spans 4 RIRs, but has no
>> customers, I also advocate changing from "and customers"
>> to "or customers", to relieve networks such as the one I work
>> for from being unfairly excluded from obtaining ARIN resources.
>>
>
> I'm ok with "technical infrastructure or customers", I've been debating
> between, and, or, and and/or myself.  Are there any objections to
> "technical infrastructure or customers"?



No objections to "technical infrastructure or customers", nope.


>
>
>  I will also note for the record that as port density increases,
>> the number of devices we use is going down, not up.
>>
>> They cost a metric shit ton more, and suck up more power
>> and need more cooling--but if you're measuring by "number
>> of boxes" rather than "capability of boxes", I think the expectation
>> that the number of boxes in a network will always be increasing,
>> as someone else further down in the thread claimed, is prima
>> facie false.
>>
>
> I don't think we want to be measuring the size of the network, at least
> the number of devices used to build the network.  Just that there is a
> network, or portion of a global network, within the region.



OK; so "20%" could be measured in terms of cost of devices in the
ARIN region, rather than device count, in the case of a network
that has a few very large, very expensive network elements in
the ARIN region, but hundreds of small, inexpensive nodes
outside of the ARIN region.



>
>
>  Matt
>>
>> (for the record, while I'm suggesting alternate language that
>> I think might be more palatable, as currently proposed,
>> I oppose this proposal)
>>
>
> Do you opposed to the whole approach?  Or, Are there changes to the text
> that would allow you to support the Draft?  Or, is there another approach
> to the problem you would propose?
>
>
 I keep feeling like we're approaching the problem from the
wrong angle.  :/  I don't have a better approach yet, but I'm
trying to noodle over it.  Fundamentally, we're trying to ensure
that entities that obtain number resources in the ARIN region
are registered with physical addresses somewhere in the
ARIN region, and that the addresses get used primarily in
the ARIN region, right?   Maybe we should stick to simpler
language like that, rather than explicitly trying to identify
percentages of customers, or costs of devices?

Thanks!

Matt



> Thanks
>
>
> --
> ==============================**==================
> David Farmer               Email: farmer at umn.edu
> Office of Information Technology
> University of Minnesota
> 2218 University Ave SE     Phone: 1-612-626-0815
> Minneapolis, MN 55414-3029  Cell: 1-612-812-9952
> ==============================**==================
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20130917/e2d8bb7d/attachment.htm>