Hello, Le 11 févr. 2015 à 21:42, Anfinsen, Ragnar a écrit :
> Hi guys. > > I am working with my management team to implement IPv6, but I got an > interesting question from one of the managers; Why do we need more IPv4 if > we are moving towards IPv6? > > A quick background; We are having discussions around IPv4 and IPv6 and the > need to eventually buy more IPv4 addresses to keep a premium level on our > Internet access. > > My argument is that we need addresses as long as there are important > services that only do IPv4 (yes, there are still a few, especially in > Norway), and as long as the other ISP are reluctant to implement IPv6 > (luckily in Norway, all the major ISPs have already come a long way). When > IPv6 reaches critical mass is the $5000 dollar question which I wish I had > the answer for. > > So, any thoughts on this topic, and any qualified guesses on when we no > longer need to do IPv4 and still be able to call our internet product > premium? Actually it depends on whether you are on the content or ISP side. But both showed a benefit. * On content side, there is an example of a french hosting company named Gandi who rents its VMs with an IPv6-only option. Benefit for the hoster : less IPv4 to find out, benefit for the client : a cheeper VM. * On ISP side, you can think about 464XLAT deployments where users may have an unfiltered IPv6 but a kind of CGN on IPv4. Benefit for the ISP : less traffic through the CGN (i never seen studies on this point but it would be really interesting), benefit for the customer : a reliable access to its favorite websites (Google, Youtube, Facebook) without the CGN factory. * On big infrastructures, you can also think about having your servers addressed IPv6-only and put an IPv4 only on your load-balancers Anyway, i think you can find a way to show a benefit according to your case. Best regards Emmanuel Thierry
