[arin-ppml] ARIN-2013-4: RIR Principles / Request for General Thoughts

[Michael Sinatra]

On 6/7/13 7:16 AM, Chris Grundemann wrote:
> Hello all,
>
> As the shepherd for ARIN-2013-4, I'd like to note that while the
> initial conversation on this draft policy was fairly constructive,
> more recent comments have gone a bit off topic. Rather than debating
> the history of internet governance it would be very helpful for the AC
> to understand the general sentiment of this community with regard to
> the draft policy more directly.
>
> With that in mind, I'd like to gather opinion and thoughts on several
> key questions. These questions should be answered in terms of how you
> believe ARIN should operate today and into the future in order to
> uphold its goal of stewardship of the Internet number resources in its
> care.
>
> 1) Do you support the principle of efficient utilization based on need
> (Conservation/Sustainability)?

Yes, but...

For IPv4, I am mainly concerned with preventing market abuse, hoarding,
bubble markets, etc., and for allowing for a smooth transition to a
market based allocation scheme in a "market" where the vast majority of
IPv4 endowments were granted for free (i.e. no cost for the addresses 
themselves, only for registration services).  I am still not sure that I 
want IPv4 to be part of a grand experiment in the monetization of 
numbers, but if we are going to go there, then let's do it carefully. 
For that reason, we still need some form of needs-basis.  But I see it 
more as loosely regulating a market rather than actually conserving 
addresses.

For IPv6 and ASNs, I am for a very liberal version of needs-basis, as
conservation of *addresses* and individual ASNs is less of a
sustainability issue.

I do not consider myself to be religiously anti-market, nor do I 
subscribe to the religion known as "free-market fundamentalism."  I am 
speaking from the perspective as a concerned network operator, who, due 
to a freak accident, has actually taken graduate-level economics courses 
at a major research university.  (I realize that last part probably 
lowers my credibility, but whatever.)

> 2) Do you support the principle of hierarchical aggregation (Routability)?

No.  BGP is not hierarchical.  Moreover, attempts to force (or at least 
encourage) hierarchical aggregation onto IPv6 by the standards community 
and RIRs has demonstrably failed and has arguably been *one* of the 
factors that has set IPv6 deployment back in terms of years.

> 3) Do you support the principle of uniqueness (Registration)?

Yes.

> 4) Do you support the goal of balancing these principles with each
> other under the overarching principle of Stewardship?

It depends on what we're stewarding.  Is the RIR supposed to be the 
steward of the number resources or of the registry itself (or both)?  Is 
the RIR the steward of the routing system?  (If so or if not, then how 
does RPKI fit in?)  This gets to the overall question of the role of the 
RIR in a world where IPv4 is basically finished from the perspective of 
the RIR allocation function, and where the relative abundance of IPv6 
(and perhaps 32-bit ASNs) leads to decreasing allocation effort, while 
RPKI and BGPSEC will demand more of the RIR.

I am not sure how important these principles are when the RIR becomes 
more of a PKI-based routing arbiter and less of an address allocator.

michael