Re: Transition to IPv6 AddressKim Liu – Aug 02, 2007 7:25 AM PST
NAT will always be required.
Presumably, at some point in the future, a successor to IPv6 will be developed. (If we are going to assume that IPv6 will be the one and only protocol for the rest of eternity, better tell that to all the students now starting to study networking in college, so they know to switch fields, as their elders have already perfected the network.)
Consider the current issues in transitioning the IPv4 network, with its address space size and the number of devices currently on it, to IPv6. Now, imagine a fully IPv6 network, without any NAT present, with IPv6 address space size, and that many more devices, and trying to transition *that* network to the protocol to the next generation.
(I am presuming a future protocol would not be backwards compatible with IPv6 due to having superior features and functionality not present in IPv6, motivating the transition. Of course, considering the dead weight and inertia in a hypothetical fully IPv6 global internet, it could very well kill the motivation for protocol development for a very long time.)
Almost certainly, some form of NAT will be needed in that situation, too.
NAT will always be required.
Presumably, at some point in the future, a successor to IPv6 will be developed. (If we are going to assume that IPv6 will be the one and only protocol for the rest of eternity, better tell that to all the students now starting to study networking in college, so they know to switch fields, as their elders have already perfected the network.)
Consider the current issues in transitioning the IPv4 network, with its address space size and the number of devices currently on it, to IPv6. Now, imagine a fully IPv6 network, without any NAT present, with IPv6 address space size, and that many more devices, and trying to transition *that* network to the protocol to the next generation.
(I am presuming a future protocol would not be backwards compatible with IPv6 due to having superior features and functionality not present in IPv6, motivating the transition. Of course, considering the dead weight and inertia in a hypothetical fully IPv6 global internet, it could very well kill the motivation for protocol development for a very long time.)
Almost certainly, some form of NAT will be needed in that situation, too.