[arin-ppml] Draft Policy 2012-4: Return to 12 Month Supply and Reset Trigger to /8 in Free Pool

[George Bonser]

" That space is just one example where consumption of v6 addresses are in decline."

Meant "v4 addresses are in decline"

> -----Original Message-----
> From: George Bonser
> Sent: Thursday, March 15, 2012 4:07 PM
> To: George Bonser; Morizot Timothy S; arin-ppml at arin.net
> Subject: RE: [arin-ppml] Draft Policy 2012-4: Return to 12 Month Supply
> and Reset Trigger to /8 in Free Pool
> 
> Another reason for the reduction in burn rate may have nothing to do
> with going from 12 month to 3 month allocations.  Think about it ...
> there SHOULD be NO change in burn rate over a 12 month period.  If I
> need a /19 every three months then I am going to need them every three
> months whether I get them once a year in a larger block or not.  The
> frequency of having to go to ARIN does not change the growth rate of a
> service requiring IP addresses.  So the notion that going from 12 month
> supply to 3 month supply having greatly reduced the consumption of
> address resources seems odd to me.
> 
> What is more likely are things like the largest consumers of public IP
> space (mobile networks, for example) going to IPv6 internally in their
> networks and using NAT64/DNS64 CGN to interface devices to the v4
> Internet.
> 
> There are an increasing number of native v6 devices out there on an
> increasing number of networks in the mobile space.  This is reducing
> the amount of v4 addresses burned up in these mobile networks.  VZW,
> for example, has a large number of native v6 devices these days and as
> old devices are retired and people upgrade to new ones, the number is
> increasing with every passing day.
> 
> That space is just one example where consumption of v6 addresses are in
> decline.  My v6 traffic is increasing with each passing month.  I would
> currently estimate that of all clients able to connect to my network
> for service, about 1% are currently native v6 (I'll know with more
> precision at the end of this month).  Of the newer clients, 100% are v6
> capable and nearly 100% are on a v6 native network with some carriers.
> 
> The reason for reduction in burn rate isn't because of the interval in
> requesting resources.  Growth is not modulated by frequency of
> contacting ARIN.  I would posit that the reason for burn rate reduction
> in IPv4 is that we are seeing more widespread adoption of v6 in
> internal networks by large consumers of address space than people
> realize.  We don't see the corresponding increase in v6 traffic on the
> internet because we don't see service providers moving their primary
> landing pages to v6 yet.  For many you still need to provide a
> "special" hostname to get their v6 service and most people aren't going
> to bother with that.  So much of the traffic is being NATed to v4 at
> the network provider's internet edge.
> 
>