2013-08-23

Co.Labs

Even The Laziest Developers Should Transition To IPv6 Now

We’re running out of IP addresses. A solution, IPv6, has been in the works for years, but too many companies and startups are relying on workarounds to delay it. Developers who embrace IPv6 sooner than later are making a better investment and may even gain a competitive edge--here’s why.



The IPv4 to IPv6 transition is like a fire drill in an office building. There are plans and backup plans and meeting points and check-ins, and during the drill hundreds of people herd down stairs and out into daylight, where they stand blinking and generally milling about, not absorbing any of it. It’s like that with IPv6, except this fire affects the whole Internet-connected world, and this is not a drill.

If you’re not privy: Internet Protocol (IP) addresses are the numbers assigned to each device on a network for identification and addressing. They specify an identity and a digital location, as with mailing something to your friend’s house at the address you looked up or he gave you. The current IP address system, IPv4, began with 4.3 billion combinations, or 4.3 x 10^9. As of this month the U.S. has less than one percent of these addresses left for allocation, and other regions of the world have similar extreme shortages. Theoretically if IP addresses ran out completely there would be no way to bring new devices online, set up new servers, or host new websites. In practice there are a number of workarounds that could extend the life of IPv4 (and are), but the idea is that it would be wasteful to invest in retrofitted infrastructure when IPv6 is ready to deploy.

But we’re not exactly being blindsided by this issue. Widespread concern about IPv4 “exhaustion” has existed among computer scientists, network specialists, and the telecommunications industry since the late 1980s, and an alternative, IPv6, has been in the works since then. It has 3.4 x 10^38 addresses or 340 trillion trillion trillion. So yeah, we’re probably gonna be able to chill with it for a while. There’s not a lot of digested and condensed information about the IPv6 transition out there, though, so I read around, watched hours of YouTube lectures (like this one) and spoke to an architect at one of the biggest IPv6-enabled internet companies in the country for some background.

You may have heard of IPv6. You may know about World IPv6 Launch, which began in June 2012. You may have realized that the network settings on your computer have included IPv6 options for a while or that Linux has supported it since the mid-'90s. You may have even tested your readiness or intentionally replaced a broken router with one that is IPv6 ready. But if you’re not on board as a developer you should be. Axel Pawlik, the Managing Director of RIPE NCC, which allocates IPv6 addresses in Europe, the Middle East, and parts of Asia, said earlier this month in TechRadar that “Businesses . . . do need to be concerned if they're investing money in hardware that will need to be updated in just a few years. They also need to be aware if they are IPv4 only, they are effectively invisible to anyone using IPv6.”

IPv6 advocates seem enthusiastic about the transition’s progression. The IPv6 World Launch Blog posted recently that “the total number of registered networks with IPv6 deployment results that we can publish [is] 216. The large increase comes about as we now have greater coverage of registered networks from more than one source of measurements, a requirement for our methodology.” 216 doesn’t sound like a big number on the scale of the Internet, does it? Furthermore, IPv6 adoption rates are about 4% in the U.S. and about 2% worldwide right now. This feels low. But the IPv6 community insists that the number will grow by leaps and bounds as large ISPs who are in the process of implementing IPv6 bring it online and begin offering it to customers.

And this is where developers really stand to gain from incorporating IPv6 into products and services now. It may seem like continuing to use IPv4 is just the quick and dirty way to get to market, but IPv4 is rapidly deteriorating in ways that can hurt new products. For example, targeted advertising and location-based services often use IP addresses as a shorthand for a user’s position. If an IP address was allocated to a certain organization with regional affiliations years ago, it can be assumed to serve a device in that region. But as single IP addresses are shared among more and more devices to minimize demand, it becomes harder to pinpoint a user’s actual location based on their IP.

Additionally, multiple users on a single IP mean that the actions of one may limit the potential to reach others. If 10 people share an IP address and one of them launches a phishing campaign that subsequently degrades the reputation of that IP, other innocent users may be unfairly penalized, and could be blocked from accessing and using your service. In a recent reflection on the IPv6 transition, network engineering consultant Chris Phillips wrote:

The opportunity to position yourself as an early adopter has come and gone, but being IPv6-ready still places you ahead of most networks. It may even create differentiation between you and your competition, giving you a leg up.

All the available IPv4 addresses for the U.S. will be allocated in the next year or two, with major ISPs continuing to dole them out for a time after that. This means it may be five years or so before hosts have no choice but to be IPv6-only. And since there are currently no plans to turn IPv4 support off, there will certainly be a long tail of consumers adopting IPv6 over time (those who replace hardware based on device lifespan will eventually upgrade without realizing it). It seems logical that ISPs won’t turn IPv4 off until the cost of providing it outweighs the economic benefit of providing it. And that will be years.

Unlike the transition to digital TV, in which the government publicized the switch and subsidized analog-to-digital converters, the move to IPv6 is simply being fueled by a long-term need. In the former case, the government and telecommunication industry had a monetary interest in the parts of the radio spectrum they would be able to reclaim after the move to digital TV, but in the case of IPv6 the impetus for change is only pragmatic. In the short term it may be tempting to “go with what you know” and stick with IPv4. But adopting IPv6 is an obvious future-proofing tactic, and there are ways to present IPv6 support to consumers as a value-added feature. So why not be bold?

[Image: Flickr user Jeff Attaway]