[ppml] Proposed Policy: Adding an HD ratio choice for new IPv4 allocations

[Charles Scott]

PPML:
  As this has been introduced before I would like to copy a message I
posted a little more than 1 year ago regarding the validity of HD ratio
with respect to ISP allocation/re-allocation. I don't believe that these
specific questions were addressed the last time HD ratio came up. I am
still not convinced that there is sufficient reason to introduce HD ratio
as it appears to address a problem that is questionably applicable to
end-user utilization and apparently not applicable to ISP
allocation/re-allocation. As the rational for HD ratio is being applied to
the needs of ISP's, I would appreciate an explanation as to how and why HD
ratio would be applicable.

Chuck Scott
cscott at gaslightmedia.com


>Date: Tue, 10 Feb 2004 08:38:01 -0500 (EST)
>From: Charles Scott <cscott at gaslightmedia.com>
>To: Michael.Dillon at radianz.com
>Cc: ppml at arin.net
>Subject: Re: [ppml] HD Ratio changes
>
>
>
>Michael:
>  When this last came up I questioned why HD Ratio should be applied to
>ISP allocations/assignments and received no responses. In fact, I even
>challenged the application of HD Ratio to end-user address space on the
>premise that the argument for HD Radio is based on the need to implement
>a strict hierarchical numbering scheme. With dynamic routing protocols it
>is not necessary to have a network hierarchically arranged strictly by
>numeric address--subnets can appear in networks to which they are not
>numerically related. The arguments for HD Ratio have used telephone
>number distribution as an example, but that is not a comparative model
>today (although number portability may convolute that).
>  Frankly I see this as being parallel to encouraging end-users to use
>NAT in that the more they use it, the more they conserve address space.
>In that respect, the more people rely on dynamic routing to permit more
>flexible distribution of address space, the easier it is to manage
>address space and the more they can conserve (and the more they avoid the
>concerns driving HD Ratio).
>  Regardless of the above, all of the references about the need for HD
>Ratio relate to end-user address assignment issues and don't seem to be
>relevant to allocating, or re-allocating (here we go with terminology
>again) blocks of addresses by ISP's. Frankly, I can't see how it would be
>relevant to that. I might be inclined to agree with the use of HD Ratio
>if it only applies to end user implementation of address space and not to
>address block assignment/allocation, and if I was convinced that the
>mitigating issues addressed above do not relieve the concerns. Otherwise,
>I'm, once again, concerned that this approach unnecessarily increases
>waste and complexity and in doing so provides questionable relief.
>
>Chuck Scott
>cscott at gaslightmedia.com