[ppml] Proposed Policy: IPv4 Countdown

[Owen DeLong]

On Mar 15, 2007, at 12:28 PM, Stephen Sprunk wrote:

> Thus spake <michael.dillon at bt.com>
>>> Imagine a day in which a lot of the network outside of the ARIN
>>> region is IPv6 and only ARIN is still using IPv4.  We'd have a
>>> network ripped in half (well, not 50/50) with our region falling
>>> behind the rest of the world (in IPvX).
>>
>> In my version of this thought experiment, the ARIN region quickly
>> realized that roughly 90% of their infrastructure was capable of
>> running IPv6 with only a software upgrade so they did upgrade.
>> Problem solved.
>
> Nice thought experiment, but 99.9999% of the routers in the ARIN  
> region (you
> know, all those boxes consumers and SoHos are sitting behind) don't  
> support
> IPv6 and the vendors have shown absolutely no interest in adding  
> it.  Not in
> their new products, and certainly not in the millions of old boxes out
> there.  And, even if it were available, how do you plan on reaching  
> out to
> millions of Joe Sixpacks and Grandmas and convince them to upgrade?
>
Frankly, you don't need to.  Almost every one of those can be  
"converted"
to IPv6 without the user even knowing it happened.  Instead of upgrading
the SOHO box, in most cases, the provider can simply 6to4 NAT them at
the DSLAM.
> An IPv6 core is irrelevant if none of the endpoints can reach it.   
> Vista
> finally has v6 on by default, and maybe 20% of users will have  
> upgraded by
> the end of this year, but they still won't be able to reach an IPv6  
> router
> even two hops away at their ISP.
>
Very true.  However, an IPv6 core is one of the steps necessary to  
resolve
that.  While an IPv6 core provides no immediate benefit in the scenario
you describe, it is even more true that an IPv6 edge will not be widely
deployed until an IPv6 core is available to connect the edge endpoints.
Nobody wanted to buy a telephone handset before they could get
a connection to some level of telephone network.

Owen