Dual stack IPv6 for IPv4 depletion

Thu Jul 16 21:21:46 UTC 2015

In message <55A812A1.6020007 at ttec.com>, Joe Maimon writes:
> 
> 
> Jacques Latour wrote:
> > Hi,
> >
> > Dual stack is where we need to go 'now', but we need to think about the fut
> ure where we run an IPv6 only stack and stop thinking how to leverage, extend
> , expand and create ugly IPv4 solutions. IPv4 is done; it served its purpose 
> well, thank you. We need a date where IPv4 is no longer routed on the Interne
> t. I am suggesting 4/4/2024. Whatever the timeline, we need to agree on the d
> ate and drive toward a common goal for a better Internet.
> >
> > My 2 Canadian cents :-)
> >
> > Jack
> >
> 
> Just as nobody is preventing you from going ipv6 only right now, I 
> advocate against hindering anybody going ipv4 only for as long as they 
> want/can.

There is nothing stopping you experimenting with class E addresses
behind a NAT.  Talk to your vendors to lift the restrictions and
route the packets as unicast packets.  Note this really doesn't
help with the global shortage.

> You may not like the results, but that does not make a top down approach 
> any better a choice.

Turning Class E into global unicast is nearly as big a project as
getting IPv6 deployed everywhere.  There is lots of equipement that
can't make the jump.  Even starting 15 years ago we would still not
be in a good faith position to hand the addresses out to be used
by anyone to talk to anyone.  Both end boxes have to support it and
all the routers in between have to support it.  It's not just about
being able to assign the address locally.  Its about everything
involed in the path being able to route to/from the addresses.

Keeping IPv4 going longer required more publically routeable IPv4
space and class E was never it.

> Joe
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org