[arin-ppml] ARIN-PPML 2014-7
Rudolph Daniel
rudi.daniel at gmail.com
Fri Feb 7 11:58:12 EST 2014
Hi
I am clear on what you are saying sir, and in 95% or more of the ARIN
region, I do not see any issues, but we have either an ixp with two
participants or you are designated an private & ISP...
I am suggesting that the stated potential for abuse of the "two" rule may
also be present in the three rule where the three rule can conceivably be
used to withhold support in some small developing states for sake of some
commercial competitive advantage.
Granted North America is looking towards a significant increase of ixp
s...so too are non North American ARIN countries who are still on a steep
deployment learning curve.
The burden may not be obscenely high, OK, but the focus for change to 3
has been set @"potential for abuse", and the scarcity of resources;
however, currently,
"This policy does not preclude exchange point operators from requesting
address space under other policies."
Rudi Daniel
(information technologist)
784 430 9235
On Thu, Feb 6, 2014 at 6:31 PM, David Farmer <farmer at umn.edu> wrote:
> There seems to be a clear potential for abuse of the two participant rule
> after run-out of the general ARIN IPv4 free pool. Also, there seems to be
> a clear potential that changing to three participants will make it more
> difficult for some IXPs to get going.
>
Agreed, and I would favor the 3 participant rule against the 2 participant
rule. It would be hard indeed, to start an Exchange later; if all the
reserved resources were allocated to 2-participant "exchanges".
The burden should not be obscenely high, for a new Exchange to sign up at
least 3 participants.
As I see it: ARIN's job after exhaustion, is to try to allocate IP address
resources required today, not to facilitate the anticipated expansion via
IXPs that would today be just expensive two-member peering arrangements
structured as a 2 member IXP in order to qualify for some extra /22 or so.
My other observation is: An Exchange with only two current participants
is not really an Exchange, but a private peering -- regardless of the
theory of possible additional participants in the future.
While it is ARIN policy not to recover addresses solely due to lack of use,
PERHAPS it should be different for IXP and critical
infrastructure/immediate need microallocations under 4.4, or 4.10.
E.g. Required notification when the number of verifiable
actively-interconnected Exchange participants drops below 2,
Or, when some or all of the microallocation is no longer being used in the
manner that justified its allocation for Critical Infrastructure:
With possible required return and renumber requirement solely due to
non-use.
Thanks.
>
--
-JH
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20140207/f8461986/attachment.htm>
More information about the ARIN-PPML
mailing list