You all remember Y2K, right? As the year 2000 approached, experts warned of the impending cataclysm as they predicted computer systems around the world would crash and burn because they couldn’t handle four-digit years. Then the fateful day came, and…nothing much happened. What was the story? The doom-and-gloomers were exaggerating the danger? Or maybe the global IT community rose to the challenge and fixed everything in the nick of time? It depends upon your point of view, even now, eleven years later.
Today we have a similar crisis on our hands. This week IANA handed out all the remaining IP addresses in the IPv4 space. Once the service providers finish passing them along to customers, there won’t be any more. At least, until IPv6 is in place, or we somehow wrest the extras out of the hands of Halliburton and the US Postal Service and the other favored organizations who received millions of IP addresses back in the early days of the Internet.
The IPv4 exhaustion crisis is eerily similar to Y2K in that the eventual outcome may go two quite different ways: either fizzle or Armageddon. But there’s also a fundamental difference: we’re not nearly as worked up about IPv4 exhaustion as we were about Y2K ahead of the fact. The question is, should we be? If you think we overreacted to Y2K, maybe not. Here are some of the issues. You be the judge.
Issue #1: Is IPv6 far enough along to support a relatively seamless transition?
YES: IPv6 has been on the radar now for about 20 years. Vendors have been building it into their hardware for at least three or four years. Your network gear probably already supports it, unless your hardware is really old. You’ll need to do some configuration changes, but no big deal.
NO: Less than 1% of the network gear in production actively supports IPv6. There’s still poor demand for such equipment, which means it’s still expensive. And simple reconfiguration? Good luck with that!
Issue #2: Will Network Address Translation (NAT) extend our supply of IPv4 addresses for long enough for IPv6 to become fully established?
YES: We’ve been doing NAT for years. Most organizations as well as multi-computer households use NAT. It’s robust, mature, and scalable.
NO: Even with NAT, we risk swamping our supply of IP addresses. Not only is every telephone, automobile, and refrigerator online these days, but our browser sessions require multiple connections to servers, which can swamp the ports necessary for NAT to work properly. No multiple connections means no Ajax means no Facebook. Welcome to the White Screen of Death.
Issue #3: Will we be able to free up enough IP addresses from the organizations who have surpluses?
YES: Many of the organizations who own the Class A address ranges are government or non-profit, and they will likely give them back for the greater good. For the ones in for-profit companies’ hands, make ‘em an offer.
NO: ZapThink seems to be the lone voice in the wilderness calling for a secondary market for these address ranges. Hello? Is anybody out there?
Issue #4: Will the open, route-around-problems nature of the Internet lead to a straightforward resolution to the IPv4 exhaustion crisis?
YES: If customers want IPv6, customers will get IPv6. IPv6 provides enough IP addresses so that every molecule in the universe can have one. So they’ll be dirt cheap.
NO: Net Neutrality is out the window. The incumbents—the big telcos who own the Internet as well as big Internet companies like Google and Yahoo!—have no motivation to make it easy to move to IPv6 for free. Instead, they’ll want to pass along the cost to the end customer. You want IPv6? Cough up the bucks.
Y2K the Sequel: June 8, 2011
You might think another difference between Y2K and IPv4 exhaustion is that Y2K had a firm, incontrovertible deadline. It’s true that there’s no one day that IPv4 goes away, thankfully, but we do have a trial run coming up: World IPv6 Day, which is June 8. On that day, Facebook, Yahoo, and Google, with a combined one billion visitors per day, are participating by enabling IPv6 for their main services that day. Content distributors like Akamai are participating as well. Fortunately, they won’t be turning their IPv4 support off, but that doesn’t mean you’ll be able to get to their sites that day either!
What will happen on World IPv6 Day? Perhaps a fizzle: people with really old browsers will have some issues, and maybe they’ll finally get with the program and upgrade. Perhaps Armageddon: billions of requests might go awry, giving most of the world the impression the Internet has stopped working altogether.
Fizzle or Armageddon? Time to place your bets, folks, the silver ball is spinning!