[arin-ppml] Fairness of banning IPv4 allocations to some categoryof organization

Robert E. Seastrom ppml at rs.seastrom.com
Thu Oct 8 18:25:05 EDT 2009


"Olivier MJ Crepin-Leblond" <ocl at gih.com> writes:

> A "Smart Grid" will require network connectivity, especially in the most
> remote places - such connectivity is only possible using IPv4 at the moment,
> thus their need for IPv4 addresses.

I don't think anyone is quite insane enough to suggest putting power
meters on the public Internet.  Yes, a smart grid will require
connectivity of some sort, and the power companies will have to be the
ones to haul it in.  At that point, the layer 3 and up protocols are
completely up to them.  They could do it with CLNS or X.25 or Chaosnet
or LU6.2 or whatever floated their boat (and suffer the consequences
associated with each).

> Or perhaps metering equipment isn't IPv6
> compatible yet?

Not for lack of trying.  The meter vendors I talked to through the end
of 2008 all understood the importance of supporting IPv6 (some of them
as a direct result of my arm-twisting).  Whether such meters are
available on the street yet, so to speak, is a good question.  On the
plus side, even under duress power generation and transmission
companies make technology moves on a geological timescale.

> I understand the concern that should such a grid be
> built on IPv4 addressing, this will accelerate IPv4 address
> depletion. I also understand it would be a lot more *desirable* if the
> power utilities used IPv6. But if they wish to implement their plan
> today, what choice do they have? How do you choose whom to ban from
> obtaining IPv4 addresses? What would the criteria be? IMHO this would
> imply a serious slowdown to innovation already.

Eh, they'll be doing it on 1918 space, at least to start, and the
first time their outsourced billing department (common, particularly
for smaller utilities) gets the chocolate and peanut butter mixed up,
they'll discover why globally unique addresses are a good idea.
Hopefully by that time the IPv4 depletion clock will have run out and
the choice will be a very easy economic one.

-r




More information about the ARIN-PPML mailing list