May 06, 2010, 4:14 PM — More Internet traffic is expected to be carried via tunnels as the Internet infrastructure migrates from IPv4, the current version of the Internet protocol, to the long-anticipated upgrade known as IPv6.
IPv4 uses 32-bit addresses and can support 4.3 billion devices connected directly to the Internet. IPv6, on the other hand, uses 128-bit addresses and supports a virtually unlimited number of devices -- 2 to the 128th power.
The regional Internet registries said in April 2010 that less than 8% of IPv4 addresses remain unallocated. The remaining IPv4 addresses are expected to run out in 2011 or 2012.
Here are the basics you need to know about the most popular tunneling and encapsulation mechanisms that are available to help transition your network from IPv4 to IPv6.
IPv6 traffic statistics
As of March 2010, only 1% of Internet traffic was based on IPv6, according to Geoff Huston, Chief Scientist at APNIC.
Huston noted that the 90% of IPv6 traffic was native, while 10% was using a tunneling mechanism to carry IPv6 traffic over IPv4 links.
Of the various tunneling mechanisms, the so-called 6to4 technique appears to be gaining in popularity, while the alternative Teredo is becoming less common, Huston says.
"The number of folks doing 6to4 as a percentage of IPv6 folks is increasingly rapidly," Huston told the audience at an Internet Society IPv6 panel in Anaheim, Calif. "The number of folks doing Teredo is really low. I'm not sure we need it."
6to4 is a method of connecting IPv6 hosts or networks to each other over an IPv4 backbone. It doesn't require explicit tunnel set-up, and instead uses relay routers to forward encapsulated IPv6 packets over IPv4 links. It uses unicast to create point-to-point links over the IPv4 backbone for transmission.
6to4 is the method of choice for users or networks that want to connect to the IPv6 Internet using an IPv4 connection. It allows these users to communicate with other 6to4 users as well as users of native IPv6 connections.
One benefit of 6to4 is that it doesn't require configured tunnels. It can be implemented in border routers without a great deal of router configuration.