[arin-ppml] Fee structures for ARIN

[Jack Bates]

On 10/28/2011 12:46 PM, Jeff Wheeler wrote:

> I don't understand how you could forecast ARIN to generate too much
> revenue from IPv6.  It should be uncommon for an organization, ISP or
> otherwise, to need more than a handful of IPv6 allocations.  On the
> other hand, we all know that IPv4 depletion has been hanging over our
> heads since before ARIN was created, thus IPv4 policy has caused all
> growing organizations to have to go back to ARIN and request more and

Handful of allocations? What has that to do with anything? Many ISPs 
will have a single allocation, no matter what the size. I don't pay by 
number of allocations for IPv4. I pay for the aggregate equivalent. And 
yes, as long as we have IPv4, ARIN does more work to maintain it. 
However, I'm not sure that actual makes up a large portion of ARIN's budget.

> more address space.  Also, I would hope networks will figure out how
> to put a lot more customers and services into an IPv6 allocation
> within a given size/fee tier than the similarly-priced IPv4
> allocations.  If not, they at least have made that choice themselves
> and there is every reason to believe they think the fees are worth the
> extra address space.
>

> If I thought people would be shutting off their IPv4 anytime soon
> because it isn't useful, I would worry about ARIN being able to
> generate *enough* revenue given the current fee system.  Perhaps that
> is a problem ARIN will have to solve ten years from now, but there
> seems to be no danger of IPv4 being deprecated in the predictable
> future.
>
Actually, with the newer policy and nibble allocations, the problem is 
that a network can grow extremely large very fast. If you need a /27 of 
IPv6, you are jumping up to a /24, which is MUCH larger. If it were more 
inline with the current IPv4 Tier system, the revenue should be near the 
same. This is why I recommended the nibble boundary pricing escalation.

ARIN should only lose revenue if ISPs get extremely tight with their 
allocations. If they follow the model and don't short change their 
customers, ARIN should be fine.

> I also do not understand why ISPs that SWIP address space to their
> customers create substantially more cost than an end-user who does not
> need to SWIP.  I know there are statistics available on the number of
> SWIP requests processed automatically vs how many require some kind of
> assistance from hostmaster at arin.net.  If ISPs can't figure out how to

They aren't. The problem is, there wouldn't be an end-user if they raise 
the end-user rates too much. The ISP is subsidizing ARIN for everyone 
else. I'm okay with this.


Jack