[ppml] IPv4 Soft Landing - Discussion and Support/Non-Support Requested

David Conrad drc at virtualized.org
Fri Oct 5 13:40:14 EDT 2007


On Oct 5, 2007, at 7:32 AM, Edward Lewis wrote:
> As an aside - if we delay the run out of IPv4 by 12 months, is  
> there an indication that the obstacle to IPv6 will be removed in  
> that 12 month period?

Unlikely.  The obstacle (IMHO) to IPv6 deployment is that it doesn't  
solve a problem anyone thinks they have.  However, with that said,  
the IPv4 Soft Landing proposal does try to go around this obstacle  
(see below).

> If it is the routing system, will 12 more months improve/strengthen  
> it?

No.  I always thought the RIRs didn't focus on routing.

> (I guess I have never understood the rationale for "rationing" the  
> remaining IPv4 addresses.  They aren't a consumable {water, oil}

An interesting assertion.  I'd argue the IPv4 _free pool_ is a  
consumable. It is the free pool that will be rationed, either  
incrementally (via increased requirements) or in a binary way  
(yesterday: yes, you can have your address space. today: nope. no  
more address space available, period.)

> and the last won't tide us over until there's a rescue.)

The point of any "soft landing"-style proposal is to attempt to buy  
more time for folks who need it to make the transition by imposing  
some form of rationing.  In my proposal the rationing is based on  
increased requirements to document ratcheted up utilization.  My  
proposal also goes a step further in that it attempts to encourage  
IPv6 deployment (given it seems the general consensus is that moving  
to IPv6 is better than increased proliferation of IPv4+NAT).

In the ideal world, the increased restrictions on obtaining IPv4  
would impose sufficient additional administrative (or other) cost  
that when taken in the context of the encouragement of IPv6, would  
reduce the barriers for ISPs to deploy IPv6 despite there being  
insufficient customer demand.  In other words, the goal is to create  
an environment that breaks the IPv6 chicken-and-egg problem.

No other proposal I am aware of attempts to do this.  To be honest,  
the two other proposals you mention feel a lot to me like rearranging  
deck chairs on the Titanic -- shuffling where the End Days free pool  
resides might encourage investment in unanticipated places but it  
does nothing to either extend the lifetime of the free pool nor  
encourage the transition to IPv6.  Neither proposal hurts anything,  
of course, rather I just don't see what advantage they bring.


More information about the ARIN-PPML mailing list