[arin-ppml] Policy Proposal: Open Access To IPv6

[Michael K. Smith]

On 5/30/09 1:47 PM, "Garry Dolley" <gdolley at arpnetworks.com> wrote:

> On Fri, May 29, 2009 at 08:57:03PM -0700, Michael K. Smith wrote:
>> Hello All:
>> 
>> I am in favor of this.  I've been following the comments in the various
>> threads and subthreads, and would add only this:
>> 
>> - There is at least one major transit provider that will accept nothing more
>> specific than a /32.
> 
> And why do you think that is?
> 
> If we give a /32 to anyone who asks for it, I *guarantee* you that
> major ISPs will stop accepting /32's.  There will be too many of
> them.
> 
> They'll take it down to /30s or /24s or something like that.

Impossible to say what will happen later.  Right now, /48's are not globally
routable over the same paths as /32's.
> 
>> - If we continue to inhibit providers' ability to get a /32, they may have
>> reachability issues.
>> - The v6 space is incredibly large.  Really.  What do we really gain from
>> limiting some people to a /48?
> 
> The number of addresses in IPv6 is incredibly large, but not the
> number of /32 subnets.
> 
> Guys, do the math.  There are the same number of /32 subnets in IPv6
> as in IPv4.  We are running out of IPv4.
> 
> We'll run out of IPv6 if there is no barrier to entry on a /32.
> 
> 2^32 = 4,294,967,296

Straw man.  4 billion /32's means 4 billion assignments to providers from
the RIR's.  This does not equate to the 4 billion addresses in the v4 pool
because they're not assigned as hosts.

Mike