[arin-ppml] Draft Policy 2010-1: Waiting List for Unmet IPv4 Requests

[Scott Leibrand]

On 1/27/2010 5:20 AM, michael.dillon at bt.com wrote:
>
>    
>> You always have great input... how do you think that 2.A. should be
>> addressed?
>>      
> The whole policy proposal is nuts.
>
> If ARIN can't give out a /14 because there is no single free block
> big enough, they should just send an email saying:
>
> ------
> Here's some good news and some bad news. We just approved your
> allocation, however, we don't have any /14 or bigger sized
> blocks left. The biggest free block available is a /16.
> We could either give you that and consider the request to be
> completed, or we could give you 3 /16s and 2 /17s to make
> up the equivalent number of IP addresses out of several
> NON-AGGREGATABLE blocks.
>
> How do you want to proceed?
> --------
>    

What if the request is for a /14, and the biggest free blocks are all 
/24s?  Do you want to give out 1024 non-aggregatable /24s to meet their 
need for a /14?  Or should they be offered a single /24 from the free 
pool, and given the option to get their /14 via transfer?  The latter is 
the outcome this policy would prefer, as it reduces fragmentation of the 
IPv4 address space, and allows available blocks to be matched with a 
larger number of equivalent-sized requests, rather than having them all 
vacuumed up by a small number of large requests.

And what about when the last /24 is given out, but there are still 
requests coming in, and there is still small amounts of address space 
being reclaimed?  Do you think that all requests should be denied if the 
pool is empty, and the first request to come in after a block is 
reclaimed gets it?  (Pool of Bethesda style?)  In that case, perhaps 
every requester could send in requests once a minute, all day every day, 
until a block becomes available.

> That doesn't need any new policy just common sense. It also doesn't
> need to waste everybody's time on PPML arguing over the last table
> scraps.
>    

Yes, the last few blocks will be small, but that doesn't mean someone 
won't want them.  IMO we need to have a mechanism in place to allow for 
an orderly distribution of the last of the free pool, and anything that 
comes in afterward.  This draft policy is one proposed mechanism for 
doing this.  I'd love to hear alternative suggestions.

-Scott