IPv4 Address Exhaustion Projections

As the more technically minded readers will know, there is a finite number of IPv4 addresses available to be allocated – specifically 2^32 or 4,294,967,296. The replacement IPv6 is not yet widely used (I do not think a single organisation in NZ is on IPv6) as transition costs are not insignificant. IPv6 will never run out I would say as it is 2^128 or 340,282,366,920,938,000,000,000,000,000,000,000,000.

Incidentially the number of stars in the universe is estimated at 5×10^22, so that is around 123 trillion IPv6 addresses for every solar system. So unless the aliens get greedy, we’re fine.

Anyway a fairly important question people ask is when will IPv4 addresses run out. An expert in this field is Geoff Huston. He knows more about this, than 99.99% of people. Huston has a script which analyses current allocations and predicts three dates.

The first is the date IANA will run out of IPv4 blocks to allocate. And that is quite soon – 16 February 2010. That’s when the global unallocated stock will run out.

The second is the date the Regional Internet Registries will run out of their local stocks, seeing they can get no more than IANA, and that is only a few months later on 21 September 2010.

Now it’s not quite the end of the IPv4 world at that stage, because there are allocated blocks which could be freed up. For example Ford Motor Company (and Haliburton!) have a Class A or /8 block of 16,777,216 addresses, and they certainly don’t use them all. The estimate is these could keep IPv4 allocations going until March 2017.

But as the IANA looks to run out in early 2010, I expect to see this being the catalyst for large scale transitions to IPv6.