The end is nigh
Enjoy it while it lasts folks, ’cause the ole Interwebs is gonna implode in just over 700 days. I’m talking about the IP address crunch. Kind of like peak oil, we’re burning though the last 10% of the available Internet addresses at a rate that puts us firmly on track to run out in mid-to-late 2011. If you take perverse pleasure from such a thing, you can watch the addresses running out before your very eyes, rather like sand though an hourglass. If that doesn’t turn your crank, Wikipedia offers a helpful executive summary of our digital plight.
“Oh NOs!”, say all the bloggers and gamers and porn site magnates. “The sky is falling! What will we do?”
Well, tadpoles, IP version 6 is supposed to be the answer — a new addressing scheme to replace the aged and venerable IP version 4. (Don’t ask what happened to version 5. I could tell you, but then I’d have to kill you.) IPv6 encompasses an address space that is so super-large that we could give out a gazillion addresses a day every day until the sun becomes a cold dark cinder and not come anywhere near running out.
“Well great!” say all the facebookers and twitter addicts. “Lets move over to that.”
Only one problem, I say (as does Dan Bernstien). When the bright boys and girls at the IETF sat down and designed IPv6, they designed it to be a replacement for IPv4 rather than an extension to it. In a nutshell, this means that there is no easy way to get IPv6 to talk to IPv4.
So, imagine it. The day finally comes in 2011 when they run out of version 4 IP’s. The next poor sod to request an address from the IANA will get, by default, a version 6 IP, and then he will discover that he can only talk to the rest of the Internet if he uses special, clunky and slow, translation services for his connection. Think Network Address Translation times ten, with horns and a pitchfork. His connectivity will suck, and it will only get worse as more and more sites are forced to adopt IPv6 and use these same translation services to reach the majority of IPv4-only sites on the Internet. Then, on top of that, imagine all the disruption that will be caused by other sites upgrading voluntarily to IPv6 and encountering all the bugs in various vendor’s implementation of the IPv6 network stack that will only come to light under sustained real-world use. Finally, imagine all the screw-ups that will happen as the Sympaticos and Verizons of the world start deploying a network stack that most of their staff have no experience with.
Changing the Internet from IPv4 to IPv6 will be rather like trying to change the tablecloth on a table that seats 500 million diners who are all eating boiling hot soup. Even if nobody gets permanently burned or maimed, it’s still gonna be a f**king mess.
But I’ll probably make a killing as an IPv6 transition consultant.
Grondzilla says:
September 18th, 2009 at 12:39 am
This brings to mind the idea that as a group we often do our best thinking in a state of crises. Perhaps as we get down to that last few months some really clever bugger is going to figure out how the transition will be managed without invoking the end times. Are there inklings of short cuts that you’re aware of that will might point in this direction?
Zilla's Other Half says:
September 18th, 2009 at 8:49 am
So the question that comes to my mind, now that the smart people who wrote IP6 have been pointed to the fact that the world will, apparently, end as a result of their hard work, is, do you think someone will raise their head out of the sand long enough in the next 700 days to find some sort of pathway that maybe will allow vs. 4 to talk to vs 6 (or vice versa)? Or are they just going to throw up their hands and say there’s nothing they can do?
desert rat says:
September 18th, 2009 at 2:44 pm
Y2K was supposed to be the end of the world, too. Far as I can tell, we’re all still here. Some super-smart nerd will figure something out before it’s too late. If not, I’ll have a great excuse to spend more time in my garden.
Mister Angry says:
September 18th, 2009 at 5:32 pm
People made a big deal about Y2K, even though nothing much happened, mostly because there was good money to be made from panicking people about it. Requesting a bigger I.T. budget for 1999? Cite Y2K as the reason. Are you an out of work COBOL programmer? Hype Y2K and convince some big bank they need you back to check over their legacy system. You get the idea. The opposite is true with IPv6 – there will be good money to be made fixing the mess once everything breaks, or to prevent everything from breaking, because the developed world really can’t function properly without the ‘net anymore. On the other hand, in these financial times, nobody wants to spend any money to prevent a problem that the majority of netizens aren’t worrying about.
Also, many decision-makers think this problem is already solved, or that it’s not a big deal, as in “We’ll just change over to IPv6 when the time comes”, so the I.T. people who are requesting funds and manpower to start changing things over now keep getting blocked by an upper management that is still thinking “oh, we’ve still got years left before the IP shortage becomes an issue, and that IPv6 thing is supposed to solve it. Besides, that’s a lot of money Bob in I.T. is asking for”
And so it goes. The money won’t get spent until people start complaining, and one thing John Q. Public has not been doing is bitching that he doesn’t have IPv6. It’s a non issue for most end users, and will continue to be so until something finally breaks, then it will suddenly become the issue. The problem is, in a nutshell, that nothing will be done until someone panics, and by the time someone panics it will be too late to migrate to IPv6 in any sane and reasonable way.
The sane migration route is to have everyone on the planet using IPv6 before the IPv4 addresses run out. Can it be done in the remaining 2 to 3 years? Probably, but it will take a lot of money and even more time. I don’t think anyone is going to find the political or financial will to do so until the big guys start losing revenue because “the Internet is broken”.