[ppml] Policy Proposal: IPv4 Soft Landing

[John Paul Morrison]

1. If you are going to try and push people on to IPv6, it makes sense to 
start at the top. ISPs, carriers, and hosting providers with the largest 
allocations will likely have more customers dependent on them, more of 
their business at stake, and will also have more in house expertise. 
They need IPv4 to do new business, so make their business dependent on 
IPv6 and they will have a motivation, as well as the means to do it.  
End users should receive IPv6 addresses bundled with new IPv4 
allocations, but end users can't even begin to drive demand unless 
there's an IPv6 backbone and some services out there.

2.  That may be the point. While there are IPv4 addresses still around, 
there's leverage. The last available IPv4 address will have the most 
leverage, but once it's gone, there's no more leverage. I think there's 
a very big assumption that IPv4 address exhaustion is the end of the 
world and that IPv6 is guaranteed to replace it in a reasonable time 
what that event does happen. Maybe there will be a huge Y2K like sense 
of urgency and effort to replace and upgrade everything, but maybe 
there'll just be more demand to tweak and fiddle with IPv4 to keep it 
running just a bit longer. And end users don't need routable IP 
addresses do they? We can NAT them all or just charge more. Why give end 
users routable and valuable IPv4 addresses when they will just use them 
for peer-to-peer copyright violations and/or trying to avoid censorship? 

So I think there's two approaches - either do nothing, and hope it all 
works out - IPv6 should win out, but I think there's a chance that IPv4 
could be around forever - there's enough invested in it and enough 
motivation to keep what's good enough working a while longer.

If ARIN is going to do something to replace IPv4 with IPv6, then there 
has to be a fixed criteria, but preferably a fixed date. With a fixed 
date, people can scramble Y2K like to make the cutoff.  If ARIN isn't 
going to do anything, then that would still be useful to spell out in 
policy - i.e. we're handing out addresses as needed, until the last IPv4 
address is gone, then it's on to IPv6 - let the market  sort out the 
transition. Or the policy might be at least be to reserve the last /8 
and suspend all existing IPv4 allocation policies at that point - ie 
require case by case, committee approval.

Rich Emmings wrote:
> Opposed as written.
>
> It a poorly conceived idea when it was "2007-12 IPv4 Countdown Policy Proposal", and the changes do not address the weaknesses in 2007-12.
>
> Comments:
>
> 1) Based on conversations, I get the impression that most folks proposing 
> implementation IPv6, _probably_ have not tried to implement IPv6 on any 
> large scale.  ping6 -I eth0 fe80::xxxxxxxxxxxxxxxx doesn't count.  (I will 
> grant I could be wrong on this, hence "probably".)
>
> 2) Artifically making a commodity rare, will cause a run on it before it's rare and push up the dates.
>
>
> In order to provide something constructive, I think these policy proposals 
> attempt to bite off too much.  Break it down into 3 separate ones.
>
> First, define events on the use curve.  (BTW, it's logistic, not exponential)
> Not only when we think we need to change policy, but why?  Maybe we only 
> define one point in time for now, and just stay focused on what we need to 
> do to not get there.
>
> Next, define different policies that ARIN can use with regard to space 
> assignment.  Whether you have a /16 or a IPv6 network might be the factors 
> that come up here, or whether it's a time based allocation -- you get a /16 
> for 2 years then have to return it, or some other agreed-to space, if 
> you're asking for some swing space while you do something else.
>
> Last, tie the first event to policies.  Once we approach that event in time, 
> we have a track record, and can discuss the next best policy to proceed 
> with.
>
> Discussing actions seperately from events, allows them to be fully discussed 
> with the ramifications, without getting into the side issues of when 
> things turn bad.  Discussing the events allows a cleaner discussion of 
> growth, live data, models and expectations.  Bringing it together last, 
> allows a discussion of what-when.
>
> If things are that rotten, then it may be that assignment policies change 
> now, even before the next event.
>
>
>
> On Fri, 11 May 2007, Member Services wrote:
>
>   
>> ARIN received the following policy proposal. In accordance with the ARIN
>> Internet Resource Policy Evaluation Process, the proposal is being
>> posted to the ARIN Public Policy Mailing List (PPML) and being placed on
>> ARIN's website.
>>
>>
>> Policy Proposal Name: IPv4 Soft Landing
>>
>> Author: David Conrad
>>
>> Proposal Version: 1.0
>>
>> Submission Date: 2007-05-02
>>
>> Proposal type: New
>>
>> Policy term: Permanent
>>     
> _______________________________________________
> This message sent to you through the ARIN Public Policy Mailing List
> (PPML at arin.net).
> Manage your mailing list subscription at:
> http://lists.arin.net/mailman/listinfo/ppml
>   

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20070514/163123d2/attachment.htm>