Re: [v4tov6transition] draft-arkko-ipv6-transition-guidelines WGLC

[Joel Jaeggli <joelja@bogus.com>]

On 8/25/10 8:25 AM, Randy Bush wrote:
>> the solution for the fixed operators is not translation 6<->4; it is
>> IPv6 deployment.
> 
> i suspect we do not disagree, but just to clarify.
> 
> fixed operators (in the non-wireless-phone sense) need to support home,
> enterprise, etc. customers who will have both ipv4 and, we hope, ipv6
> needs for a while.  so we need to deploy ipv6 *along side* ipv4, i.e.
> dual stack.  stateless translation tools do not scale or perform well at
> isp backbone scale.
> 
> our dual stack edge customers will use our dual stack backbones in a
> simple fashion, no impedance matching needed.  the ipv6-only customer
> will need nat64/dns64 to reach legacy ipv4-only destinations.
> 
> the ipv4-only customer will use ______ to reach ipv6-only destinations?

prayer.

you're back to the typical problem, you make a query, you get back only
a aaaa record, so you're dead.

since the likelyhood of you changing the v4 only customer's stack is
limited you're stuck with alg's, and presumably only for a limited
subset of applications.

> randy
>