[ppml] Policy Proposal 2007-23: End Policy for IANA IPv4 allocations to RIRs - Revised Text

[michael.dillon at bt.com]

> This policy describes the process for the allocation of the remaining
> IPv4 space from IANA to the RIRs. When a minimum amount of 
> available space is reached, one /8 will be allocated from 
> IANA to each RIR, replacing the current IPv4 allocation policy.

This is a deeply unfair policy. It means that when we near the point
of exhaustion of IPv4, IANA will give some people a 3 year supply
and some people a 3 month supply.

It would be much fairer to reduce the size of IANA allocations from
a full /8 to a smaller block size. In this way IANA can ensure that
everyone runs out of IPv4 addresses at the same time.

I suggest that the block size be based on the RIR whose yearly
allocation of IPv4 addresses is the smallest, and be some defined
fraction of that RIR's yearly fraction. For instance, if the 
5 RIRs have the following annual allocation rates:

A. /7
B. /8
C. /16
D. /10
E. /9

Then the smallest RIR issues a /16 in one year. For instance we
could say that when there are no more than 5 full /8 blocks left
in the IANA free pool (an easy and unambiguous measure) that
we would change the IANA allocation size from /8 to x times 
the minimal of the 5 RIR yearly allocation rates. In this case
it would be x times /16. We could either define x as a constant
or have some formula to define x. If we decide that x should be 1,
the the IANA allocation size shifts from /8 to /16.

This does not mean that an RIR cannot receive more than a single
/16 from IANA. It means that we move to a system more like the
RIR allocations to ISPs/LIRs. If RIR A can show that it needs
5 /16s to handle the next half-year of allocations then that is what
they should receive.

Blocks would be adjacent to previous ones whenever possible.

But 2007-23 is bad policy, and inherently unfair. I strongly oppose it.

--Michael Dillon