[arin-ppml] ARIN-prop-154 Shared Space for IPv4 Address Extension (w/IETF considerations) - Staff Comments

Owen DeLong owen at delong.com
Tue Jul 5 20:05:30 EDT 2011

On Jul 5, 2011, at 3:51 PM, Joe Maimon wrote:

> Owen DeLong wrote:
>> The IAB advised that allocating the space would be contrary to RFC 2860.
>> They did not address the issue of identifying, reserving it and setting it
>> aside without allocating it which is what proposal 154 specifies.
> Tell me how an identified reservation contrasts positively with an actual allocation. You risk a de-facto allocation with none of the benefits of an actual allocation.

The proposal requires ARIN to identify it, but, not to disclose its identity as I
read it. An identified reservation that is not published contrasts with an actual
allocation in exactly the fact that it is unpublished.

I'm not sure where the risk you claim would come from if the reservation is
not published.

Identify, to me, in the context of proposal 154 distinguishes between "we need
to stop allocating when the free pool has only a two /10s in it" vs. "this particular /10
is specifically reserved for this purpose". Nothing says that the identity of the
particular /10 needs to be published by ARIN.

>> I am concerned that there may or may not be a /10 block available by the time
> I believe this should be addressed with an unidentifiable reservation.

I am not sure what the point would be to simply tell ARIN to hold off
on allocating space beyond a certain limit vs. asking them to set aside
a particular /10. It is the publication of the number range that seems
to be your issue, not the identification of it. Since there is benefit to this
being a contiguous /10 rather than the other reservation policy
(NRPM 4.10) which requires only space which in aggregate represents
the equivalent of a /10, I think it is better to reserve a particular /10
for this purpose vs. 4.10 which allows ARIN to leave the determination
of which pieces comprise the reservation to a later point.


More information about the ARIN-PPML mailing list