[ppml] Policy Proposal: Resource Review Process

[Stephen Sprunk]

Thus spake <michael.dillon at bt.com>
>> The nature of the "review" is to be of the same form as is
>> currently done when an organization requests new resources,
>> i.e. the documentation required and standards should be the same.
>
> This text is not in the policy itself but is in the explanation. As
> such, it is not binding on ARIN.

There's no binding text that says what the current review process is, so 
there's nothing to reference or quote.  We went over this last time, and my 
answer is the same now: if you don't like the NRPM's complete lack of 
guidance on what "justify", "documentation", etc. mean, submit a proposal to 
fix it.

> But it does raise the question. If the nature of the review is the same
> as is done for new allocations, then what is the point of having a
> review policy at all?  Are there that many organizations that do not
> apply for new allocations every year or two?

LIRs are supposed to be getting a six-month supply of addresses.  The point 
of this proposal is to address folks who _don't_ apply for new 
allocations/assignments "every year or two", which means mostly folks that 
got way more than they needed.  There may be some inconvenience to folks 
that got exactly what they need and aren't growing, but isn't it within 
ARIN's charter as a responsible steward to find out that's the case?  (And 
then, presumably, not bother them again for a long time)

> Just how many addresses in total are allocated to such organizations?

I have no data on the number of /24 equivalents, but the majority of direct 
assignments fall into this category, as do a minority of LIR allocations.

> Or is the entire point of this policy to slip in the language which
> gives legacy address holders extra rights above and beyond the
> majority of us.

Through lack of policy action to change historical practices, legacy holders 
_already_ have extra privileges.

> Let's look at point 8:
>
> 8.  Legacy resources in active use, regardless of utilization, are not
> subject to revocation by ARIN.  However, the utilization of legacy
> resources shall be considered during a review to assess overall
> compliance.
>
> If an organization is a legacy address holder and also has non-legacy
> addresses, then they could lose the non-legacy addresses if their
> overall usage is not sufficient.

That is correct.

> But they cannot lose any of their legacy addresses regardless of
> what their usage level is. If a company has a Class B /16 legacy
> block in which they can only justfy one /24, then ARIN can *NOT*
> recover the additional 255 /24's.

That is already true today; this just codifies existing treatment of legacy 
space.

> This is not a fair and balanced policy.

The first and second times around, this policy was opposed in part because 
it wasn't clear if it affected legacy space.  We're going to get flak on 
this area no matter what we say until the community manages to get consensus 
on what to do about legacy space in general.

S

Stephen Sprunk      "Those people who think they know everything
CCIE #3723         are a great annoyance to those of us who do."
K5SSS                                             --Isaac Asimov