Next Net moves forward

The next generation of the Internet, known as Internet Protocol version 6, took another big step toward commercialization, as its second phase of testing in North America wrapped up last week.

Moonv6, the network used in the testing, is an IPv6 backbone built by the University of New Hampshire InterOperability Laboratory (IOL), the North American IPv6 Task Force, Internet2, the U.S. Department of Defense and more than 30 networking vendors, testing vendors and service providers. The virtual IPv6 backbone stretches from New Hampshire to California and will be permanent, so that equipment makers, software vendors, service providers and anyone else who would like to test IPv6 interoperability in a live network can do so.

For two weeks in March, the North American IPv6 Task Force, the Defense Information Systems Agency's Joint Interoperability Testing Command and others tested the network for quality of service, security, application handling, networking protocols and end-to-end domain name server functionality on all major operating systems.

IPv6 expands the pool of unique addresses available for connecting PCs and other devices in the Internet. It is widely regarded as a necessary successor to the current system for Internet addressing, IPv4, which many people say does not provide enough space in its address field to support the millions of devices that will likely be added to the Internet in the next several years.

Mobile communications and new IP services, such as voice over Internet Protocol and video on demand, will increase the number of devices that need IP addresses. The IP address shortage is likely to affect Asia and Europe first, where adoption of these new technologies is growing fast. Few analysts expect the problem to impact U.S. networks anytime soon.

"We are seeing interest in Asia, particularly in Japan," said Rose Klimovich, vice president of global IP virtual private networks at AT&T. "And in the U.S., we're seeing interest mainly from the government. There have been a few customers interested in IPv6, and they are starting to talk to us about it now."

Klimovich said AT&T's network is ready today for IPv6, but customers haven't yet asked for the service. NTT, which also participated in the testing, offers an IPv6 service in the United States today. France Telecom, another participating service provider, doesn't currently offer an IPv6 service, but it can deliver some IPv6 traffic to the United States.

The Defense Department is the main driver behind the Moonv6 project and the adoption of IPv6 in the United States. In June 2003, the department set a mandate that all agencies be IPv6-ready by 2008. As a result, the Moonv6 project was launched to provide a test bed to certify equipment and work out protocol kinks.

The first phase of testing was completed in October 2003. This initial phase only tested basic routing applications and simple network configurations. The second phase was much more complex and advanced, said Ben Schultz, who works at the University of New Hampshire's IOL. It included more robust testing of routing protocols, such as Border Gateway Protocol and Open Shortest Path First. It also tested firewall functionality and quality of service, as well as 10-Gigabit Ethernet.

The Moonv6 project showed that IPv6 could be used on the same network as IPv4. This aspect of the test demonstrated that the two protocols can work in parallel, which should help ease the transition to the new standard.

Now that testing is complete, the network will remain live for informal testing. Organizers plan to continue growing the network by adding new service providers and network nodes.

Technology companies that participated in Moonv6 included AT&T, France Telecom, Agilent Technologies, Check Point Software Technologies, Cisco Systems, Extreme Networks, Foundry Networks, Fujitsu, Hewlett-Packard, Hitachi, Lucent Technologies, Microsoft, NEC, NetScreen Technologies, Nokia, Panasonic, Procket Networks, Spirent, Sun Microsystems and Symantec.

[InetUser]

Don't need IPV6 for address space

I just want to remind everyone that some are trumpeting as the real reason to move from IPV4 to IPV6: increase in address space.

This is complete bunk. If the IPV4 addresses are more effectively managed, they would go much farther. You have to look at those that abused the early address pool by grabbing an entire, or even multiple class A network blocks.

Break up some of these huge blocks and reallocate them, and the existing address pool would suffice for a large number of years.

OK, so why do they want IPV6? Because in IPV6's design are multiple bug-a-boos that make it really easy to track and filter every packet, which service(s) it's destined for, where it came from and going to and a bunch more invasive information collecting. It's probably the largest privacy hack around. That's why.

Not to mention that you'll probably have to replace nearly ever router that has to deal with IPV6, so CISCO and the other mfgs are pushing for IPV6.

So I say, let's not go IPV6. Let's make the IPV4 address pool last as long as we can instead.