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

[Kevin Kargel]

One thing to remember that may affect your algorythm, is that for
residential ISP's as the last mile deliverer we have two edges..  an
external edge connecting to the world, and an internal edge connecting
to our customer premises.  

Right now the only solution that makes any sense for me is to dual stack
both edges, then dual stack my core, slowly migrate my customer edge to
v6 only, and migrate the outside edge ot v6 when v4 traffic disappears
(indicating end user need dissapation).  

I certainly believe we are running out of v4 addresses.  I do not
anticipate that v4 will be obsoleted in my career duration.

Kevin


> -----Original Message-----
> From: David Williamson [mailto:dlw+arin at tellme.com] 
> Sent: Thursday, February 14, 2008 11:20 AM
> To: Kevin Kargel
> Cc: ppml at arin.net
> Subject: Random v6 discussions (was Re: [ppml] Policy 
> Proposal: IPv4 Transfer Policy Proposal)
> 
> On Thu, Feb 14, 2008 at 08:49:52AM -0600, Kevin Kargel wrote:
> > 	As customer connections are the primary consumer of 
> IP's for an ISP, 
> > utilizing IPv6 for this will allow IP's to be reclaimed for use for 
> > global connections.
> 
> That's one of the key problems, isn't it?  The place you 
> least want to disrupt anything is the very edge.  We're 
> talking about mucking up layer 3.  That's the network 
> layer...that bit that's supposed to provide end-to-end 
> connectivity.  Everything below isn't worried about reaching 
> beyond the local network.  Everything above is assuming the 
> network layer has done its job.  You can muck about in the 
> middle of a connection (by encapsulation, labelling, etc.), 
> but the end points are going to care about that end-to-end 
> connectivity.
> 
> 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.  That implies some quality time 
> with protocol translators...pick your favorite 4-to-6 
> conversion method.
> 
> It stuns me that there are serious networking folks who don't 
> think we'll run out of v4 addresses.  But this is now 
> waaaaaay off the topic of policy, which is the purpose of this list.
> 
> -David
>