On 6 Aug 2019, at 9:05 am, Mark Tinka <mark.tinka@seacom.mu> wrote:
On 2/Aug/19 14:17, Baldur Norddahl wrote:
The pricing on IPv4 is now at USD 20/address so I am thinking we are forced to go the CGN route going forward. Of all the options, MAP-E appears to be the most elegant. Just add/remove some more headers on a packet and route it as normal. No need to invest in anything as our core routers can already do that. No worries about scale.
Actually, I think NAT64/DNS64/464XLAT is the best option, because as more IPv4 falls away, you are automatically translating less and going native IPv6 more. And there is nothing for you to "turn off" or migrate away from after all is said & done.
Mark.
Which only applies to DNS64 and not 464XLAT. That said, every IPv6 node should be attempting to connect over IPv6 first. That alone moves most of the traffic to IPv6 regardless of the IPv4aaS method in use. DNS64 also breaks DNSSEC which is not a good thing. DNS64 alone also depends on *everybody* having good (complete) IPv6 connectivity and not leaving IPv6 breakages uncorrected. There is no fallback to IPv4 with DNS64 alone. If you also have 464XLAT with DNS64 then there is NO DIFFERENCE to MAP-[ET] or DS-Lite in terms of traffic shifting to native IPv6. Mark -- Mark Andrews, ISC 1 Seymour St., Dundas Valley, NSW 2117, Australia PHONE: +61 2 9871 4742 INTERNET: marka@isc.org