[ppml] And as for assignments...

William Herrin arin-contact at dirtside.com
Sat Aug 25 20:33:23 EDT 2007


On 8/25/07, Jonathan Barker <jonathan at qx.net> wrote:
> Assigning a /64 to a home... or 2^64th addresses... which is the number
> of IPv4 addreses available on the Internet today - SQUARED... Surely I'm
> not the only person who thinks that's just crazy.

You're not the only one. A /64 is 65000 times the number of addresses
than all of the globally unique ethernet MAC addresses ever assigned.
You can assign an IP address to the fridge and every piece of food in
it and still have an insane amount left over.

The idea as I understood it was some sort of attempt to fix the PI/PA
problem by making the final 64 bits static and the first 64 bits
dynamic. Those final 64 bits would always come up the same no matter
what subnet the IPv6 host was plugged into. It was actually a pretty
clever idea, but it doesn't appear to have worked in the sense that it
doesn't actually solve the PI/PA problem. Or maybe it was intended to
eliminate ARP but it doesn't manage to do that either. You still have
to advertise your MAC to IP mapping since folks can still manually
assign explicit IPv6 addresses to an interface.

The criminal part, though, is it consumes so many bits that it
forecloses the possibility of solutions to PI like
http://bill.herrin.us/network/ermpi6.html . Its waste on a scale that
surpasses even the IPv4 multicast mistake.

Regards,
Bill Herrin


-- 
William D. Herrin                  herrin at dirtside.com  bill at herrin.us
3005 Crane Dr.                        Web: <http://bill.herrin.us/>
Falls Church, VA 22042-3004



More information about the ARIN-PPML mailing list