Re: Reality (was RE: Stupid NAT tricks and how to stop them.)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 10-apr-2006, at 16:35, Noel Chiappa wrote:

Many years ago now, a funny thing happened on the way to "complete exhaustion of the IPv4 address space (Version 1)". Some clever people worked out this ugly hack, which the marketplace judged - despite its ugliness - to be a superior solution to the forklift upgrade to IPv6. It's been selling like
hot-cakes ever since, while IPv6 languished.

That's the popular view. In reality, people deployed NAT mostly for reasons that have little to do with the global IPv4 address depletion. And IPv6 hasn't been ready for any kind of deployment until the early 2000s.

I've become rather disenchanted with my crystal ball, which seems quite
cloudy of late (if you'd told me, in 1986, we'd still be running a
Destination-Vector routing architecture for a routing table of this size 20
years later, I'd have *known* you were bonkers),

The future just doesn't want to honor the principle of least astonishment: what we expect to change, often stays the same, while what we expect to stay the same, more often than not changes.

Don't be surprised if the world, facing "complete exhaustion of the IPv4 address space (Version 2)" decides, yet again, that some sort of Plan B is a
better choice than a conversion to IPv6.

Everyone who thinks that regular users are going to forego IPv4 connectivity in favor of IPv6 connectivity as long as IPv4 still works to a remotely usable degree is a card carrying member of the Internet Fantasy Task Force*.

For now, the usability of IPv4 is relatively constant while that of IPv6 is much lower, but steadily increasing over time as IPv6 support in hard- and software increases in quantity and quality. Assuming that the people who get all those millions of IPv4 addresses every year actually use them for something, like connecting new customers, in 4 to 8 years, something will have to give. I don't think the big change will be in the demand side, as we see that countries with several IPv4 addresses per capita (even without legacy /8s) are still using up new ones while other countries have a lot of catching up to do, and more IP devices seems likely for just VoIP if nothing else.

(I've made a new page that shows addresses per capita: http:// www.bgpexpert.com/addressespercountry.php There are actually several countries that have a factor 1000 fewer IPv4 addresses per capita than the US and a factor 10 is fairly common even in Europe.)

I have no idea exactly what it will be (maybe a free market in IPv4
addresses, plus layered NAT's, to name just one possibility), but there are a lot of clever people out there, and *once events force them to turn their attention to this particular alligator*, don't be surprised if they don't
come up with yet another workaround.

Well, the IETF has done its job by creating IPv6, so whatever happens, we should be in decent shape. Soon enough we can turn our attention to the fact that we're still doing our interdomain routing with RIP on steroids. (-:

Iljitsch

* coined by Tony Hain



_______________________________________________

Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf

[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]