Lightly used IP addresses
drc at virtualized.org
Sat Aug 14 03:05:00 UTC 2010
On Aug 13, 2010, at 2:51 PM, Nathan Eisenberg wrote:
>> I'm not against ARIN, I think they have good intentions. I'd like to think so anyway.
> Same here. I'm honestly surprised that there is as much dissention from this attitude as there seems to be...
I suspect the issue arises when ARIN (or anyone else for that matter) attempts to assert dominion over resources folks consider their own. That is, in the original scenario John Levine posed:
"1. A sells a /20 of IPv4 space to B for, say, $5,000
2. A tells ARIN to transfer the chunk to B
3. ARIN says no, B hasn't shown that they need it
4. A and B say screw it, and B announces the space anyway"
I believe the point of contention lies in step 3. In the case of the 38% of the address space described by John Curran as "managed" by ARIN with an (L)RSA, there is contractual language that dictates ARIN has some authority to "say no". In the remaining 62% of the space (according to ARIN), presumably space allocated without any form of RSA, the issue is, at least to my mind, far less clear.
In the face of this lack of clarity, when you have folks saying things like:
"6. ARIN receives a fraud/abuse complaint that A's space is being used by B.
7. ARIN discovers that A is no longer using the space in accordance with their RSA
8. ARIN reclaims the space and A and B are left to figure out who owes what to whom."
without proviso about whether an (L)RSA is applicable, it isn't particularly surprising that folks who can imagine themselves as (or at least sympathize with) A or B getting their dander up.
In addition, as/after the IPv4 free pool is exhausted, there are going to be lots of folks who discover they have more address space than they really need as there are going to be lots of folks who are desperately in need of additional IPv4 addresses. This will result in address markets. Some people (e.g., I'm guessing Vadim) do not see a role for ARIN as mediator of an exchange between these two sets of folks. Others believe that there needs to be some 'regulator' of the market or (e.g.) speculators will swoop in and buy up all the allocated-but-unused IPv4 address space, resulting in those in desperate need of IPv4 addresses paying through the nose. Given the arguments between free vs. regulated markets generates much heat in pretty much every other economic discussion, I'd be surprised if it didn't occur in address markets.
More information about the NANOG