Why hasn't the market developed?
Because there is no business justification. No one is in any danger of running out of IP addresses. People have learned to use Network Address Translation (NAT), where you mask your internal IP addresses by only externally expressing the address of the NAT router. It not only helps conserve IP addresses, but also adds security by hiding the IP addresses of internal devices.
There are two schools of thought: one is that traffic should travel directly from end to end, and the other is to use NAT and wall off subnets from the public Internet. With NAT there is no need to go to IPv6 for a long time. Is IPv6 being adopted in the U.S.?
Here, there is no IPv6 market. You only really hear about it during demonstrations of IPv6-compatible products. The one place where there will be demand for IPv6 is in the military. The Department of Defense has mandated that, by 2005, any new network gear be IPv6 compatible. I think they will be turning it on by 2008. But the military has a unique network. It is not a typical business environment. In a battlefield with remote sensors, you need a lot of IP addresses, and that unique environment actually justifies IPv6. Are there any dramatic benefits to IPv6 that may compel a business to make the leap?
IPv6 has IPsec [Internet Protocol Security] built in, so people used to talk about security [improvements in IPv6]. But that is not much of an issue anymore. Companies are already using IPsec or Secure Socket Layer (SSL) VPNs. There are not that many advantages, unless you are willing to get rid of NAT, which might simplify addressing in the network. But I don't know anyone who is anxious to do that. Companies like to hide their addresses behind firewalls so they cannot be attacked.
A lot of what we are seeing with IPv6 is a typical attempt on the part of equipment vendors and service providers to differentiate themselves from each other. IPv6 helps them stand apart a bit, but no one is asking them who really needs it. What about network management and monitoring?
There are tools for managing and monitoring IPv6. The things I'm not seeing are the tools for IP address administration -- tools to keep track of which IP addresses are associated with which subnets. In order to support IPv6, how must businesses change their networks?
Over time, companies will migrate to IPv6-compatible switches, routers and operating systems in PCs and servers. As long as a company takes its sweet time, there may not be much of a hardware and software cost. The real cost is in administration and management. It is a much more complex network environment. You need to have all these overlays on the physical infrastructure, which complicates troubleshooting.
There will also be a transition period where you are running both IPv4 and IPv6, and that can be difficult to manage. Dual stacks of devices complicate network administration and troubleshooting. What might help drive adoption?
The only thing that might help drive it is the military. In the mid-1980s, there was a new protocol called TCP/IP [Transmission Control Protocol/Internet Protocol]. No on expected to see it go anywhere. But the military began using it for battlefield communications. Ultimately, that technology became pervasive.
FOR MORE INFORMATION:
What are the technical differences between IPv4 and IPv6? Our expert has the answer.
A layer three Ethernet switch does packet forwarding in the hardware. The problem with IPv6 packets is that they are so much larger that they can't be handled in the hardware. Typically, the way that switches and routers handle IPv6 is in the software rather than the hardware, which is the slow path. One of the things that customers need to do is make sure that vendor products support IPv6 in the hardware layer, otherwise it can be an order of magnitude slower.