[ppml] IPv6>>32

Tony Hain alh-ietf at tndh.net
Wed May 11 18:22:43 EDT 2005

Howard, W. Lee wrote:
> ...
> I think that argument is a little misleading.  Extrapolating
> (32 bits / 30 years) = (64 bits / 60 years) assumes that past
> performance is a predictor of future allocation [1].
> The problem, as I like to frame it, is that if we take the
> attitude that IPv6 is infinite, then addresses or subnets will
> be assigned to consumables, consumed, and never reused.  I see
> lots of engineers-cum-science-fiction-writers who talk about
> potential uses for IPv6, all of which seem to me to argue against
> themselves; they seem to say there will be such demand that we
> can't afford to be careful now.  Especially since produce-truck
> drivers, nanobots, and complex molecules, are not ISPs as
> currently defined by policy.  If these clusters are going to be
> sliding gracefully from network to network with their /48s, then
> the /32 aggregation goal is blown.

No it is not. They are not moving the upper 48 bits, they are fitting
gracefully into the new provider's existing /32 aggregate. This is not about
consumption without reuse, it is about creating a mechanism that prevents
artificial lock-in due to the pain of rebuilding the entire network.
Readdressing IPv6 hosts is trivial by design. Renumbering network components
requires much more work in places that currently require manual
intervention. At the end of the day though as long as the subnet topology
stays the same we are talking about a string replace on configuration files
from the old provider's /48 to the new one. 

> The magic of a /64 is that it's a single routable entity.  If I
> assume that layer 3 networks connect layer 2 networks, I still
> haven't seen any argument here about what a layer 2 network of
> 2^64 devices would look like.  It's not only inconceivable, it's
> inconceivable to (2^32) power or more, and then we say that we
> have to assign this enormous set of numbers in groups of 2^16.

The point is consistency for the end site. 

> The /48 rule applied equally to every possible definition of
> end user is arbitrary and reckless.  I will concede that in
> IPv6 terms, a /48 is reasonable for the enterprise network,
> but I just can't understand how it's more responsible to assign
> this much to every node on the theory that in someone's wildest
> dreams, any node could someday be a router.  If a node does
> someday become a router, is renumbering one device (possibly to
> a /48) insurmountable?  Even if it is, is there any barrier to
> the household (or cellular, or vehicular) router routing (or
> switching) /96 subnets within the /64?  Presumably, my blinds
> will never have a host address (EUI-64, MAC, or whatever) that
> conflicts with my vacuum cleaner.
> Oh, for the record, Geoff Huston's model said we'd consume a /1
> to a /4 in 60 years; at the rate of one /1 every 60 years, we
> run out of IPv6 space in 120 years.

And also by his model if we simply change the HD from .8 to .94 that becomes
1200 years. By all means the right thing to do is have a reasonable HD

> Lee
> [1] OK, technically it assumes an exponential progression, but a
> predictable exponent.  And I don't think most people see that.

What most people seem to miss is that if 2^32 == 30, then with no other
changes 2^45 == 245,760 years. Other than Geoff & David's presentation on
why the HD metric is wrong, I have yet to see a valid argument on why 45
bits is not enough for the foreseeable lifetime of any protocol. 


More information about the ARIN-PPML mailing list