IPv6 PI allocation

[Roger Jorgensen]

On Fri, 18 May 2007, Colm MacCarthaigh wrote:
> On Fri, May 18, 2007 at 07:53:28AM +0200, Roger J?rgensen wrote:
>> On fre, mai 18, 2007 01:40, Colm MacCarthaigh wrote:
>> That leave us with 32-3-(32-29) = 26... the highest amount of network
>> available are something around 2^26. It's not _that_ many really.
>
> That's higher than my number! I was more conservative and said 2^21.
> Even 2^26 is a effectively an AS per 5 people on the planet, even if the
> population quadruples within the lifetime of IPv6, that's one per 20
> people on the planet. As I said, it's not a bet I'd take, I like
> overprovisioning by orders of magnitude, but that should really bring
> home how ridiculously big the number space is. Oh, and even if we
> totally mess up, we've burned a /3, there's a few more where that came
> from.

that was a really quick calculation from my side, quite sure I forgot a 
-2 or -3 somewhere. The number you had sound a bit more correct, but 
bottom line is, the number of IP-addresses in total is huge, but the 
amount of usable IP netblocks ain´t that high.


-- 

------------------------------
Roger Jorgensen              | - ROJO9-RIPE  - RJ85P-NORID
roger at jorgensen.no           | - IPv6 is The Key!
-------------------------------------------------------