General

IPv6 Off to a Slow Start

When it comes to IPv6 support, U.S. and Euro-ISPs are putting off today what will continue to put off tomorrow. But Japan has made IPv6 adoption a mandate—putting Asia on the front lines of the push for IPv6.

by Jim Thompson [June 12, 2001]

A soft economy, the slow rollout of 3G wireless and a who cares attitude have some wondering if IPv6 will get off the ground anytime soon.

Despite a boost from Cisco Systems, which recently announced support for IPv6 in most of its routers, the technology still can't seem to get any respect. The biggest stumbling block is a feeling that if it ain't broke, why fix it?

"The general feeling is not to be too concerned about the shortage of IP addresses now because we will all be retired when it happens, so why worry about it now" said David Preston, vice president and general manager of InterNetShare.com.

When it was first announced, IPv6 seemed to be the answer to everyone's problems. It promises a number of improvements over the current standards—IPv4—in several areas including security, management and addressing.

Some even claim that IPv6 would help create what they call a personalized Web for everyone on earth. These experts claim that if each device connected to the Internet had its own permanent address, service providers could customize the look and features of the Internet to the individual user.

"The idea is to remain connected wherever you are to whomever you choose," said Al Javed, CTO for Nortel Networks at the recent IPv6 Summit in Ottawa. "We have a good opportunity to personalize services in the wireless Internet context and service providers can make more profits through personalization, matching users with the services they want."

Theoretical benefits
But until IPv6 is actually implemented, the benefits are only theoretical. Much of the edge has been taken off the technology because many of the security and management benefits have already been tacked onto IPv4.

As a result, some claim that only one major advantage to IPv6 remains—a dramatic increase in the available number of IP addresses. IPv6 would expand the current address space from 32 to 128-bits, resulting in a nearly unlimited number of IP addresses.

This is no small matter since many predict that the four billion addresses available under IPv4 will be depleted by the year 2005 brought on by an anticipated explosion in the number of devices, like cell phones, PDAs and even kitchen appliances, that will need their own address.

Available addresses are getting short, but add-ons to IPv4, primarily Network Address Translation (NAT) have, so far, been able to deal with the problem. NAT enables a LAN (local area network) to use one set of IP addresses for internal traffic and another set for external traffic, allowing a single true IP address to link to thousands of private addresses.

Although assigning addresses and configuring devices would be much simpler under IPv6, this has not been enough of an incentive to force the widespread deployment of the new protocol—at least, not in North America where available IP addresses are plentiful. More than 70-percent of the world's IP addresses are assigned to North America. Europe has 17-percent and Asia/Pacific has only nine percent.

With so many addresses available and patches to IPv4 already in place, there appears to be no hurry to adopt the new protocol in U.S. and Canada. That doesn't mean it won't happen, just that it may take more time than originally anticipated.

3G wireless lag
3G wireless, the one technology that many thought would give IPv6 the big push it needs to get off the ground, is also lagging behind predictions.

Companies who want to play in the wireless space have spent huge amounts of money to obtain 3G licenses. At a recent auction in the U.K. more than $30 billion was spent by only five companies on licenses. Some of these successful bidders have since found themselves strapped with massive debt and not enough liquid capital to begin building the networks.

Ironically, The biggest advantage of 3G is also it's biggest problem—it's always on. This raises the question of how to bill customers once the networks are in place.

With 2G, it was simple, you just billed a customer for the time online. With 3G, a method has to be developed to charge for the type of service used and the amount of bandwidth consumed. This means developing and deploying a new monitoring infrastructure—a prospect that's not only difficult but also costly.

As 3G stalls, ISPs, at least those here in the U.S., are finding that they don't need as many IP addresses as they once thought. Those same ISPs are also finding it far more cost effective to continue using their old IPv4 networks instead of making a switch to IPv6.

"What we don't have to drive IPv6 is the killer app," said Preston. "We had the fear, uncertainty and doubt of running out of V4 address, but we don't yet have the application that only works in V6. That's the catalyst that's missing."

Another problem that IPv6 is fighting is the cost and difficulty of its deployment. In order for it to work, all network servers and workstations will have to support IPv6. Currently, now Sun Solaris 8 has native support for IPv6, but most other OSs require patches that are not always perfect. Microsoft Windows won't support the protocol until the release of Windows XP.

Even when operating systems like Windows incorporate IPv6, it will likely be three to four years before it becomes a standard in the enterprise.

Asian influence
Because of their large population and shortage of IP addresses, Asia will be at the forefront of the push for IPv6. The Japanese government has mandated the incorporation of IPv6 and set a deadline of 2005 to upgrade existing systems. Japan sees IPv6 as a way of helping them leverage the Internet to rejuvenate the Japanese economy.

Nippon Telegraph and Telephone (NTT) launched the first native IPv6 backbone in late April of this year. Of course, NTT owns Verio, a U.S. based Web hosting company and backbone operator, and has strong alliances with AT&T. It seems clear that they hope to demonstrate the benefits of IPv6 and then migrate it to the U.S. market.

The shortage of IP addresses is already being felt by countries like China and Russia who came late to the Internet game. They're trying to make up for lost ground by bringing their populations online, only to be faced with a critical lack of addresses.

In order for the power and potential of IPv6 to be realized, it must have widespread deployment. This means that every ISP, network provider, and major enterprises must fully support the protocol. This is a costly proposition, which comes at a time when companies are having a hard time justifying the expenditures.

It also does not appear that the U.S. or Canadian governments are going to try to push the technology through mandates or tax incentives. Judging from past performance, neither government appears eager to get involved in that level of technology policy. They are more likely to leave such decisions up to the private sector.

"By and large, the government is out of the Internet business, and that's a good thing," said Preston. "The Japanese are taking it on as a national pride to be a world technology leader. Unfortunately, this just doesn't work in our culture. I don't see this being mandated in the U.S."

IPv6 is alive and well. It's also not going away. But it's beginning to appear that it could take a lot more time than originally thought before it becomes the standard.

—End