[ppml] Random v6 discussions (was Re: Policy Proposal: IPv4 Transfer Policy Proposal)

[Iljitsch van Beijnum]

On 14 feb 2008, at 18:19, David Williamson wrote:

> Much as I like the idea of "fix the edge so that we have more time to
> fix the middle", I think it's going to be much more practical to fix
> the middle, and dual-stack the edge until v6 is viable end-to-end on
> its own.  Oh, and there's no way that's going to happen without  
> massive
> impact to the whole network, since getting the edge fixed will
> definitely take longer than the remaining time before pool exhaustion,
> no matter how you calculate it.

So let's not waste any time... It's still almost impossible to buy a  
broadband modem / home router / CPE that will do IPv6, and because  
those almost always do NAT, it's also pretty hard to tunnel IPv6  
through such a box.

But that's not the only hard part. ISPs can pretty much leave old  
customers on IPv4 and give IPv6 to new customers. For content sites,  
it's different: you do v6 or you don't. Because of firewalling and  
less than optimal routing in some places, IPv6 can be worse than IPv4,  
so the way things are now, it's not a good idea for Big Content to  
turn on IPv6. They also don't care about the IPv4 depletion, they only  
need a few addresses. ISPs on the other hand use up millions. So it's  
likely that we'll end up in a situation where as of a certain date, a  
lot of new users will be IPv6-only or IPv6+not-so-good-IPv4, while  
existing users and content are pretty much IPv4-only.

> That implies some quality time with
> protocol translators...pick your favorite 4-to-6 conversion method.

Within the IETF there are currently discussions about this. This is  
mostly in the requirements gathering stage, so if you have any, please  
let us know. However, being the impatient type, I co-authored a  
proposal for modifying the NAT-PT transition mechanism so that it  
doesn't have the downsides the existing way of doing that has, and it  
can also support IPv4-only applications:

http://www.muada.com/drafts/draft-van-beijnum-v6ops-mnat-pt-00.txt

(Internet-Drafts submission seems to have problems, I get a 404 from  
the IETF server right now, so please use the link above.)

The next step after this is a way for home gateways to translate IPv6  
to IPv4 to provide translated IPv4 connectivity to IPv4-only systems  
in a home that only has IPv6-connectivity.

Any feedback is appreciated, even from those of you who don't have the  
time to read the full proposal.

> It stuns me that there are serious networking folks who don't think
> we'll run out of v4 addresses.

Running out of addresses is like running out of toothpaste: there's  
never a point when you're actually flat out, but at some point you get  
tired of squeezing the tube harder and harder.