[ppml] Address Space versus Routing Slots

[Michel Py]

Hi Geoff,

<OT>
I might have a few leftover cans of black-colored paint suitable for
aeronautical use that is rumored to absorb radar signals and make the
aircraft unidentifiable; need any to re-coat your chopper? :-D
</OT>


> Geoff Huston wrote:
> What appears to be obvious (well obvious to me at any rate)
> is that at the time when an RIR can no longer meet allocation
> demands via provision of unallocated address space (i.e. the
> RIR "runs out") then the current policy framework also reaches
> its current end point.

Existing policies should be kept even then IMHO, because a few blocks
might be freed over time. Possibly some kind of a waiting list with no
guarantees whatsoever might be implemented. Without it, what would a RIR
do if space became available? I don't think much of it would happen
though, but it might be wise to plan for it.

Besides, I think there still will be need for a policy framework for the
continuing (renewed) use of already allocated space.


> Exhaustion of the IPv4 unallocated address pool does not imply
> complete unavailability of IPv4 address resources to industry
> players. i.e. the exhaustion of the unallocated IPv4 address
> pool does not appear to imply a forced IPv6 conversion onto the
> industry at that point in time.

Probably true in the US and quite possible in some other industrialized
countries who tend to have large (>1) IP-per-capita ratios.


> There is reason to believe that the Internet industry will
> continue to use IPv4 as a base protocol well after this IPv4
> unallocated address pool exhaustion date comes and goes.

Ack.


> a) market-related considerations
> Is the emergence of such markets Good or Bad?

Define "good" and "bad" please :-D


> Avoidable or Inevitable?

IMHO some of it is inevitable.


> Appropriate or Inappropriate?

What difference does it make?


> Fair or Unfair?

Life is not fair nor markets.


> Would such markets be regulated? How? What
> is the RIR role in such an environment?

IMHO the RIRs should not try to regulate prices or who has priority to
get the IP blocks that are for "sale". I do think though that the RIRs
should keep maintaining the WHOIS information. I do not like the idea of
IP blocks being sold, I much prefer the RIRs continuing to lease them. 

Therefore I envision that when time has come RIR policies should allow
the transfer of IP blocks (even if the transfer is not the by-product of
a merger or acquisition, and even if there is money involved) and make
sure that the identity of the new recipient is known, that they will pay
the annual fees, and that they will abide by policies in place.
 
The rationale behind this is as follows: if the RIRs try to regulate
prices or who can "buy" or "sell", it will inevitably create a black
market. The bottom line is that if there is demand, there will be
supply.

In the USA, we have an interesting historical prospective about this,
called the prohibition. In their greatest righteous stupidity, our
legislators managed to pass a constitutional amendment (which makes an
act of congress look easy) to make alcohol illegal. As it turned out,
the good people of the United States of America wanted to continue
drinking booze, and all the prohibition achieved was to line the pockets
of the mafia and a well-known political family (not because of their
alleged ties to the mafia but because they cleverly bought rights to
import booze during the prohibition for pennies and cashed in when it
ended).

As the Internet is not regulated by a single country or entity, it
appears to me that trying to prevent or aggressively regulate the trade
of IPv4 addresses after there are no more to allocate is doomed to fail.
What the RIRs should do IMHO is to make sure that such trade stays on
the table and does not move under the table.


> In the area of RIR Allocation Policies, there are the policy-related
> questions of: What is the threshold point where the application of
> different IPv4 address allocation policies may be appropriate? Or is
> "no change" a wiser course of action?

I would favor a "no change" course of action until we get within 18
months of the predicted exhaustion (from a RIR prospective, not IANA),
at which point the policies to allow transfer of IP blocks shall be
discussed, and implemented immediately after the RIR allocates the last
bit.

> Or should the RIRs establish "strategic reserve address pools"? Why?

The problem about strategic reserves is defining a policy about who gets
a piece of them based on some perceived need. This easily becomes an
ugly political nightmare that will inevitably create inequalities (why
him but not me). I would tend to say this: as long as there's food
everyone can feast, when there's no food left everyone starves.

I'm not trying to say that strategic reserves are a bad idea or
discourage them though, the points I'm trying to make is that a)
reaching consensus on any will be difficult and b) the cure might be
worse than the disease. I think strategic reserves are better discussed
around a specific proposal than in the absolute.

Michel.