[arin-discuss] Status of realigning the IPv6 fee structure?

[Luke S. Crawford]

On Wed, Mar 14, 2012 at 02:27:15PM -0400, Jean-Francois Laforest wrote:
> I agree with Mike too. We are expanding on the commercial side of the 
> business, considerably, but these costs would be hammering our clients 
> with a very high cost for IPs. Therefore reducing our ipv6 
> footprint/efforts due to costs. Client adoption of the technology 
> requires sane pricing.

I agree.  In my very small corner of the market, the business people
don't care at all about IPv6.  "Maybe later."   The Engineers want it, 
and eh, if they are my customers, I throw it in for free.  But as far 
as I can tell, the business types don't see a business case for IPv6
other than keeping their nerds happy.  

I mean, sure, I'd pay three grand a year for a v6 allocation, no problem;
I spend that much on stuff I evaluate and end up throwing out, even at my
size. 
 
but I only have an actual business case for this because my core customer 
base is the nerds that wanna play with this.  (And for them?  there is a 
very rational business model.   At some point, a spike in IPv6 interest 
is very, very likely, and at that point, having a few years of experience 
under your belt likely will make you rather more valuable.)  Even for
the nerds, it's a second-tier network.  If v6 breaks it's not nearly
the hair on fire big deal that a broken v4 is.  

I only mention this because yeah, three grand a year isn't a lot
for even a small company, but I think anything more than 'comes free with
the IPv4'  is going to get a thumbs down from most of the business people
I know.  

Personally, I think you want to make it a whole lot easier to get IPv6
allocations.   Right now, most IPv6 installations (again, at my very 
small corner of the market.)  are still in the 'the nerds are playing 
with it' stage, and the business people, generally speaking, wish they'd
spend less time on it.   E.g. the best way to enable this is to let
the nerds go ahead without making them talk to the business people
much, which means IPv6 needs to come free with IPv4 or something else the
business people already approved.  

I mean, at least temporarily, I think it might make sense to let people 
get IPv6 PI space before they can get IPv4 PI space.   If your IPv6 space 
causes you less headaches than your IPv4 space (and PI space causes 
vastly fewer headaches than PA space)  that's a pretty valid reason
for the business people to start moving more stuff over.   

Of course, that needs to be balanced against the routing table cost of
more PI blocks, which is why I say 'temporarily'  or something.  (I 
mean, obviously you can't take these things back, but you could say
something like the first X v6 PI blocks are at the reduced price, 
after which the rules revert to normal.) 

But my main point is that from where I stand, the business people don't
care much at all about IPv6, so you are best off slipping it in for free
whenever anyone gets any IPv4.   Certainly for the foreseeable future,
there isn't any problem giving everyone with a IPv4 PI block of any size
a /32 rather than a /40 or whatever.  The limit on IPv6 is going to 
be routing table size, right?  not address space?  so as far as our real
resource constraints go, a /40 of v6 PI space "costs" just as much as a 
/32 of PI space, so if we are going to give out a pi block, why not give 
people a /32?




-- 
Luke S. Crawford
http://prgmr.com/xen/         -   Hosting for the technically adept
http://nostarch.com/xen.htm   -   We don't assume you are stupid.