Network Computingwww.networkcomputing.com

The recent flurry of announcements from Cisco, Verio and other vendors, along with a show of support from the U.S. Department of Defense, would lead you to believe we are about to give birth to the IPv6 generation.

Sure, IPv6 will be a beautiful baby, with elegant solutions for address allocation, security, QoS (Quality of Service), hierarchical routing and limitless IP addresses. U.S. carriers and service providers are already shipping IPv6 products, mainly to the Pacific Rim, especially Japan, where denser populations demand IPv6's address space.

Don't start boiling water just yet, though. Moving to IPv6 in the next couple of years would be premature, since it will be many years before we exhaust the current pool of IPv4 addresses.

Indeed, a lot of the problems IPv6 solves have already been addressed. Although many consider NAT (network address translation) a kludge, it has eased pressure on the IPv4 address pool and provided a modicum of client security through obfuscation.

We recommend a cautious approach, letting early adopters work out software bugs and figure out ways to communicate with the rest of the world. Most of us have plenty of other things to do before the water breaks.

Post a comment or question on this story.