[arin-discuss] IPv6 as justification for IPv4?

[Owen DeLong]

>> Recognize that there is a transaction cost involved in issuing (or 
>> transfering) an address block to a party, but beyond that point,
>> ARIN's actual costs are not significantly different for a large IP
>> address block versus a small IP address block.  We should all be
>> very thankful for this, as ARIN's costs would have become enormous
>> upon the assignment of the first IPv4 block...  (which has so many
>> individual IP addresses that any cost per IP would still be too much.)
> 
> Well I think that depends and is highly subjective. In my experience the time ARIN spends on a ticket directly correlates to the size of the requested block. With the larger block I have found ARIN spends exponentially more time vetting the documentation, diagrams, spreadsheets, projections, and contracts than it does with say a /22. From what I'm hearing you say you believe ARIN spends an equal amount of time vetting a /22 as it does a /14. I can't fathom how this could be possible. If I'm requesting a /14 ARIN would presumably be reviewing documentation for hundreds of thousands of IPs and huge diagrams and projections vs reviewing documentation for 500 IPs associated with a /22 request.
> 

Nope.

ARIN spends a lot more time on a poorly considered, poorly documented repeated rounds of asking for additional documentation request for a /22 than they do on a well considered, well documented request for a /14. This has been
stated in various forms multiple times, so it does not "lack foundation" as you are so fond of saying.

Further, even in the case of a well formed request for a /22 and a well formed request for a /14, ARIN does not spend anywhere near 256 times as long on the /14 as the /22, yet you want to jack the price up *256 for that spread. In my
experience, it's more like 1.5-2 times as long. Admittedly, my knowledge is limited to IPv4 from /24 to /12 and IPv6 from /48 to /24. I don't have experience applying for anything larger than /12 (IPv4) or /24 (IPv6). My experience does include multiple successful applications at each of those top sizes.

ARIN should be spending considerably less time deliberating most IPv6 requests than they do most IPv4 requests, since the policy is quite a bit simpler and allows for significantly more liberal allocations. Also a larger fraction of these are likely initial allocations with near automatic qualification.

Owen