In an IPv6 future, how will you solve IPv4 connectivity?

Brian E Carpenter brian.e.carpenter at
Sun Oct 10 22:22:09 CEST 2010

On 2010-10-11 03:05, Roger Wiklund wrote:
> Let's say for arguments sake that the prophecy is true, and in late
> 2011/2012 a new user can only get an IPv6 address.
> Have you guys concidered/tested how you will solve these users
> connectivity to the IPv4 Internet?
> I guess NAT-PT is out of the picture,
> NAT64?
> DS-Lite?

Without contradicting the other replies, let me try this version of
an answer.

1. IPv6-only connectivity does not exclude having a dual stack host
that reaches IPv4 via an IPv4-in-IPv6 tunnel. Since there are no more
IPv4 addresses, that means using an RFC1918 address and a NAT on the path,
but that is hardly a new user experience. ds-lite is one way to do that.

2. Or an application proxy, which will work fine for HTTP and email,
and the user experience won't change noticeably.

3. Or a NAT64. Again, the user experience is just like NAT.

All of these are real solutions, not slideware.

However, I think any content provider should avoid all of these issues
by moving to a dual stack model as soon as possible. Why cause your
users the hassle of any of the above three solutions?

> Also, as these new users are IPv6 only, how can IPv4 hosts communicate
> with them? 4to6 NAT?

If they are clients, why will this arise except for p2p applications?
They are going to have to deal with this, for sure, but the p2p
community is pretty ingenious. I would think they'd do it by
proxying whereever they find a dual stack host/supernode.

If they are IPv6-only servers, I suppose we will just invert the
above scenarios. But would you invest in a content provider whose
business model was IPv6-only? ... No, I didn't think so.


More information about the ipv6-ops mailing list