[arin-ppml] /20 initial allocation for single-homed server?

[Ted Mittelstaedt]

On 5/24/2010 4:35 PM, Michael Loftis wrote:
>
>
> --On Monday, May 24, 2010 6:27 PM -0500 James Hess <mysidia at gmail.com>
> wrote:
>
>> On Mon, May 24, 2010 at 3:29 PM, Ted Mittelstaedt <tedm at ipinc.net> wrote:
>>>
>> [snip]
>>
>> Well, I would suggest then that technical justifications have some
>> artificial restrictions imposed, above and beyond policy, so that...
>>
>> (1) Additional IP addresses 'needed' solely to evade IP-based blocks,
>> blacklists, or rate limits should be rejected

The problem is (IMHO) we have NO mechanism in the NRPM to express
this as a directive to ARIN.  While Michael's comments are I believe
directed at the second of your comments, (and I agree with the
unworkability) they really don't cover the first.

Your comment highlights what I think is a reasoning we are going to
see sooner or later.  It's going to go something like this:  We
are out of IPv4 numbers, but there's got to be a lot of IP numbers
tied up by people who don't need them or are using them for
fraudulent purposes, and it's going to be cheaper to go after
these addresses than to switch to IPv6 right now.  With the
"right now" being kind of a moving target that keeps moving into
the future.

One of the responses to this reasoning is 2008-7 which WAS
adopted last year - but STILL has not been completely implemented,
see:

https://www.arin.net/policy/proposals/2008_7.html

ARIN really needs to complete implementation.  They have had
almost a year to complete it.  With this implemented the address
block allocations will be cleaned up and it will be a lot easier
for ARIN to defend against this kind of reasoning.  I think we
probably would all agree that IPv6 implementation would not be
served by ARIN being forced into a crash program to find these
alleged "available" IPv4 resources.  Thus by getting this implemented
we are assured of the integrity of WHOIS and thus people can
see for themselves that there isn't a huge amount of available
IPv4 out there.

Unless of course, there really IS a vastly more amount of IPv4
out there and ARIN is deliberately sandbagging on the 2008-7
implementation because of this.

Ted