On Wed, Feb 18, 2015 at 5:39 AM, Anfinsen, Ragnar < [email protected]> wrote:
> We are deploying IPv6 (soon) and we are not buying IPv4 for postponing > IPv6 rollout. Obviously, if buying IPv4 addresses costs less and is higher quality than something like MAP-E, then it makes sense to buy addresses and go dual-stack instead of going IPv6-only. I'm wondering what will change that equation in the future, industry-wide. Do we expect that future equipment have MAP-E built in, and thus that the technology to do MAP-E inline simply becomes available at zero cost as hardware refreshes? Or do we expect that IPv4 addresses will increase in price until it becomes a bad idea to keep buying? Somehow I get the feeling that it won't be "IPv4 traffic goes down close to zero" that gets people to move to IPv6-only. Ragnar, what do you expect will get your network to move IPv6-only eventually? You likely won't still be running native IPv4 in 2030. How will you get there?
