Waiting, Waiting and Waiting for IPv6
Page 1 of 1
IPv6, as an Internet numbering specification, is ready for prime time deployment around the world. But it seems that two key essential elements that could trigger large scale the U.S. and the Internet Engineering Task Force aren't ready to switch from IPv4.
The U.S., flush with IPv4 addresses, and the IETF, which is the standards-based body responsible for designing a smooth transition to the new numbering scheme, have been slow to adopt IPv6 specifications until the details are smoothed out.
An IP address, in its simplest form, works just like a telephone number. To contact your friend, you have to dial their exact phone number in order to speak with you pal. With an IP address, the only difference is that you use a computer or another Internet-enabled device to connect with a specific, numeric destination. Each device has a unique 32-bit identifier that labels both the sender and receiver to make a connection over the Internet.
The problem is, the number of IPv4 addresses are dwindling. Because the U.S. snatched up much of the available naming space, the address squeeze is being felt mainly in Asia and Europe. The U.S. operates the most IP addresses in the worldroughly 3,012,735,145more than enough for all to be content with the IPv4 numbering specifications.
In other words, the U.S. has 22 percent of the population of China and South Korea combined, and China and South Korea have two percent of name space that U.S interests own.
Running on empty
The demand for additional naming space overseas comes mainly from the rapid growth of Internet-enabled mobile devices like PDAs and 3G digital wireless phones. To access the Internet, each device needs an IP address to distinguish itself.
Enter IPv6, a technology that uses a 128-bit identifier to tag the sender and receiver. In addition to the exponential increase in IP addresses to accommodate everyone around the world, the new IP addressing scheme allows for more tailored communications.
IPv4 addresses, with its four octets of 8-bit blocks (for example 22.214.171.124) to identify and route Internet traffic, couldn't adequately pack in the information to provide secure transmissions, mark data packets as a priority or provide security tagssomething IPv6, with its 16 octets, can accomplish.
Kathryn Korostoff, President of Sage Research, Inc., said there is a lot of debate over the time the world will actually run out of IP addresses.
"There's a lot of uncertainty about how quickly we will run out, which is why some people have a sort of sense of panic," she said. "The whole premise is that the proliferation of mobile devices like laptops and PDAs and smart phones is going to cause us to run out of IP addresses."
She predicts that under current conditions, the world will run out of IP addresses under IPv4 in the next two or three years, as digital wireless phones and PDAs gain in popularity throughout the U.S. Several countries, including China and India, have already run out of IPv4 space.
Engineers have been working for more than two years to find a solution to migrate and integrate IPv4 and IPv6, since the two don't naturally mesh together. Migration tools have been proposed by scientists around the world to find a fix.
The devil inside
The IETF is responsible for building a path for the smooth migration of IPv4 specifications to IPv6. Back in March, however, the organization put a halt on the development of all IPv6 migration tools so it could get a handle on the technology before a slew of standards caused problems.
While the technology is stable today, IPv6 isn't quite ready for worldwide adoption because of all the migration tools required, at least according to Margaret Wasserman, IETF next-generation and IPv6 transition working groups co-chair, and chief technologist for Wind River Systems, Inc.
"Technically, we're ready today," Wasserman said. "The core standards are pretty much all addressed."
But the devil, as they say, is in the details. The core standards have a multitude of tools used to move the packets from an IPv4 to IPv6 environment and back again.
Dual stacking, the popular standard which operates in both IP addressing schemes, has more than eight tools that need to be approved, while the other standard, IP tunneling, has more than five.
The IETF needs to test all the tools before they will give their seal of approval, and called the development halt so they could at least address the tools already made.
The IETF has stated it would open up IPv6 for tool development this summer and expects worldwide IPv6 adoption sometime in the next couple years. Critics say the development shutdown stifles innovation and delays an IP addressing problem that's already come to a head. China and India have already run out of IP addresses and are migrating to IPv6, with or without worldwide approval.
But Korostoff doesn't think the rush to IPv6 should come at the cost of skimping on the details.
"There's a tactical benefit of the IETF wanting people to be practical, until things are a little more settled it doesn't make a lot of sense to have all these tools proliferating," she said. "I don't think it's necessarily a terrible thing."
The importance of being earnest
To get the U.S. to adopt IPv6 sooner would require much more incentive. Unlike China and India, the U.S. doesn't have the same incentive to switch to an unproved technology.
"In the U.S., I've seen in the past six months a reduction in the urgency of deploying IPv6," Korostoff said. She thinks the only way the county would put some urgency behind the move to IPv6 is when the crunch in IP space affects U.S. businesses and its customers.
"A lot of people look at this and think the demand is coming from consumers who want more e-mail addresses," Korostoff said. "But we're definitely seeing more businesses using mobile devices in their work force.
"It's the carrot and stick scenariodo this or else you will run out of IP addresses, that's the stick," she added. "But there's also a couple carrots. With IPv6 there's a lot more room in the protocol for information about how you handle that [data] packet."