[arin-ppml] Draft Policy 2011-5: Shared Transition Space for IPv4 Address Extension - IAB comment

[William Herrin]

On Sat, Jun 25, 2011 at 11:11 PM, John Curran <jcurran at arin.net> wrote:
> > In keeping with the spirit of RFC 2860 with respect to the assignment
> > of specialized address blocks, ARIN Staff will consult with the IANA and
> > the IAB regarding implementation of this draft policy.
>
> > == A Procedural Issue: ARINA 2011-005 and RFC2860 section 4.3 ==
> >
> > The IAB honors and values the division of responsibilities as
> > documented in RFC 2860 section 4.3. That section forms the basis for
> > Unicast address allocation via ICANN through the RIR system.
> >
> > Policy proposal 2011-005 is not a regular proposal in the sense that
> > it adheres to Unicast space. In contrast, it allows for an allocation
> > of addresses for special and global use very similar to, and almost
> > indistinguishable from, RFC1918 local addresses. Because of the impact
> > beyond the ARIN region the management (i.e. creation
> > and subsequent changes) of such reservation should be global and RFC2860
> > puts the management responsibility with the IETF.
> >
> > The IAB believes that the adoption by ARIN would be in conflict with the
> > provisions in RFC2860 and would set a bad precedent: Setting aside
> > special addresses should be done within the existing process, i.e. by
> > the IETF.
> >
> > If there is consensus for 2011-005 in the ARIN region we would be
> > happy to work with you to resubmit the proposal to the IETF and, as
> > usual, have the IESG judge consensus. This would include our reaching
> > out to other RIRs to have members of their community provide input on
> > this proposal. Clear support from the various RIR communities might
> > bring new insights into to the IETF, producing a level of support that
> > was not present with the earlier drafts.

Hi Folks,

In light of the IAB's objection, it seems to me that the ARIN board
has four options to consider:

1. Submit an internet draft as the IAB requests, along with the
implications of doing so.

2. Implement 2011-5  as recommended by the AC and community, and over
the IAB's objection.

3. Abandon 2011-5. Proponents may make their case to the IETF.

4. Implement 2011-5 as a temporary stopgap policy pending further IETF action.



1. The case for complying with the IAB's request:

The Internet standards process works because of the cordial and
cooperative atmosphere between the various NGOs and individual
participants. The IETF is indeed the appropriate venue for global
assignment of IP addresses to specific purposes as opposed to specific
end users. However, we must observe that the IANA has insufficient
unicast (class A, B or C) addresses available to award the IETF for
the implementation of 2011-5's intended use.

Accordingly, ARIN should reserve the /10 that 2011-5 calls for and
hold it unused while championing an RFC through the IETF's standards
process that uses the /10 as contemplated in 2011-5. Upon publication
of such an RFC, the /10 would be ceded to IANA for use with the RFC.
Upon a failure to reach consensus within the IETF process, the /10
would be returned to the ARIN free pool for general use.


2. The case for implementing 2011-5 over the IAB's objection.

a. ARIN's constituents have expressed a well defined, well supported
and consensus need for addresses to be used similar like RFC1918 space
but with the expectation that such space crosses the administrative
boundary between ISP and end-user and should, thus, not be used by the
end-user.

b. The IETF had the opportunity to act on these constituents' concerns
but failed to take leadership citing, among other reasons, that it
would deplete the pool of addresses available to the RIRs from IANA.

c. The ARIN region is satisfied with depleting its own address pool
for this purpose.

d. The IAB's suggestion that the proposal be brought back to the IETF
is rendered disingenuous by the fact that no addresses remain at the
IANA for implementation.

e. Precedent exists for RIRs to unilaterally act on regional
imperatives despite potentially global impact. Witness APNIC's
abandonment of needs-based allocation.

Because of these points, ARIN has a moral duty to act on behalf of its
constituents. Should the IETF desire to reclaim leadership in this
matter, ARIN's open public policy process is available to all comers
who may request that the addresses be reassigned to any RFC that is
produced.


3. The case for abandoning 2011-5:

The IETF is the proper venue for a proposal like 2011-5. Such
proposals were considered and rejected. 2011-5 is an end-run around
around the proper process.


4. The case for 2011-5 as a stopgap pending IETF action:

ARIN constituents within the ARIN region have an immediate and
pressing need for IP addresses to be used for the interior of multiple
NAT translators. This need is not adequately served by the delay
inherent in initiating a fresh proposal in the IETF's standards
process, it is not adequately served by RFC1918 and it is poorly
served by having every ISP use its own unique space allocated by ARIN.
Due to the IETF's failure to act in a timely manner while addresses
were still available to them from IANA, ARIN has a duty to act on its
constituents' imperative.

Nevertheless, global assignment of addresses to purposes rather than
registrants properly belongs with the IETF. ARIN should facilitate the
IETF retaking the leadership on the matter by ending the ARIN-region
policy and ceding the /10 address block back to IANA after the IETF
debates, drafts and publishes an RFC that the board of trustees
believes meets or exceeds ARIN constituents' expectations for these
addresses.


For your consideration,
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