Re: [sig-policy] Inter-registry transfers (was:thoughts on prop-068)
On 16/02/2009, at 10:51 PM, Izumi Okutani wrote:
Hi all,
As a related point, we had a bit of discussions here in JP over
whether
to allow an Inter-RIR/NIR transfer and opinions were generally
favourable towards allowing inter-RIR/NIR transfers.
i.e. We support this element of the proposal(s).
I have two clarifications about inter-registry transfers:
1. Size of minimum transfer
How would the minimum size of transfer apply for inter-RIR
transfers?
Read the proposals that says the source and the recipient follow
policies of respective regions - so would the policy of the region
with shorter prefix apply?
e.g. APNIC accountholders (min:/22 or /24) --> RIPE (min:/21)
the minimum size of transfer = /21
Hi Izumi,
thanks for your report on the discussions in JP on these topics.
Yes, you are correct in that interpretation of my intent as the author
of this policy proposal. In this example case the prefix size needs to
meet the apnic minimum size criteria AND the RIPE minimum size
criteria so that a transfer in either direction that spans RIPE and
APNIC would need to be a minimum size of a /21 if this is the RIPE
policy.
2. Allowing NIR-APNIC transfers
Could we suppose transfers between APNIC-NIR(at least JPNIC) account
holders can be accomodated even if there is no consensus on
inter-RIR transfers? (i.e. prop-068)
We strongly hope it can since NIR account holders are no different
from others in the APNIC region.
I suppose that in terms of strict formalism the answer would be "no,
that would be an incorrect supposition" from the strictly limited
perspective of this policy proposal. The proposal does not have
optional components that can be adjudged by the community
independently as to general consensus in acceptance. But of course the
policy process is one where proposals are put before the community in
an attempt to find what would gain such general acceptance, and if
proposal 68 fails and there is an identified need to address the
specific issues relating to transfers that encompasses members of NIRs
and members of APNIC then further policy proposals would doubtless
appear that would address that specific issue independently of the
inter-RIR topic.
Of course in all this flurry of hypothetical future policy proposals
there is the time element lurking behind all this, and I'm not sure
that the general economic downturn has really altered the basic
dynamics of IPv4 address consumption all that much in terms of gaining
extra years to debate the issue (check out http://www.potaroo.net/tools/ipv4/predict.png
to see the change in predicted exhaustion dates over time), so at
some point the inevitable will occur and the IPv4 address allocation
system used today will come to a natural halt. We probably need to
find some acceptable answers to these issues this year as to what we
do afterwards. (Unless of course we crave the added excitement of
living right on the edge! :-))
Kind regards,
Geoff