[ppml] IPv6 incentive policy

Owen DeLong owen at delong.com
Wed Aug 22 01:33:02 EDT 2007

On Aug 21, 2007, at 10:08 PM, Scott Leibrand wrote:

> Are you sure #9 is the most important?  I thought it was #10.  :-)
Take your pick.  Unfortunatley, I hear that Mr. Goldberg will be  
unavailable to present
at the meeting, due to a most unfortunate lack of pulse, so, I may be  
forced to present
this if the AC moves it forward.

> I'm also interested to know what the units of D are.  You said the  
> "number of delays".  I think you meant days, but in that case D  
> would always be 1 until C got larger than F.  Perhaps you meant for  
> D to be measured in years, such that if F is 100x as big as C,  
> you'd have a delay of .01 years (3.65 days)?
Yes... D is in days.  I suppose some scalar could be multiplied  by D  
in order
to cause these effects to be more pronounced early on and grow faster.

I think that 400 might be a good starting number such that C/F=0.01  
result in 4 days, but, C/F=.5 would produce 20 days.

Admittedly, I didn't spend much time thinking about fine-tuning the math
prior to posting this.  I do, however, think that if we fine tune the  
value to be applied to D, we will end up with a result which comes  
close to what Mr. Anderson has proposed in a much less computationally
intensive and much more predictable delay-producing policy.

So... Assuming we can find a scalar value, would you support or oppose
this policy?

Modifications included inline below.

Thanks for your suggestions.


> -Scott
> Owen DeLong wrote:
>> The recent discussion of IPv4 rationing got me to thinking, and,   
>> somehow,
>> this is the conclusion I came to.  I will point out that item 9  
>> is  probably the
>> most significant information in the template.
>> 1.	Policy Proposal Name: IPV6 incentive through IPv4 processing  
>> delays
>> 2.	Author
>> 	a.	name: Owen DeLong
>> 	b.	email: owen at delong.com
>> 	c.	telephone: 408-921-6984
>> 	d.	organization: DELONG
3.	Proposal Version: 1.1
>> 4.	Submission Date: August 21, 2007
>> 5.	Proposal type: new
>> 6.	Policy term: temporary
>> 7.	Policy statement:
>> 	In order to encourage organizations to use IPv6 rather than IPv4,  
>> 	will introduce a delay in processing new IPv4 applications which is
		computed such that D=X*C/F where D is the number of delays which
>> 	ARIN will wait to process the IPv4 template, C is the number of IP
>> 	addresses currently in use by the applicant, and, F is the number of
>> 	IP addresses currently in the IPv4 free pool (computed as the sum of
>> 	the IP addresses in the ARIN free pool and the addresses remaining
		in the IANA free pool). X shall be defined as _____. (See rationale
>> 	ARIN may completely process the IPv4 template and reserve the
>> 	addresses, collect fees, etc. immediately upon receiving the  
>> template
>> 	using it's normal process.  However, a delay shall be introduced
>> 	which shall start on the date of the final data submission by the
>> 	applicant prior to approval, and, which shall extend D days such
>> 	that the value of D is computed as defined in the preceding
>> 	paragraph.  If the value of D is computed to anything less than
>> 	one, then, it shall be treated as if the value was 1.
>> 	IPv6 templates shall be processed immediately without any delay.
>> 	This policy shall expire when the IPv4 free pool becomes double
>> 	the size of the IPv4 free pool as of April 1, 2008.
>> 8.	Rationale:
>> 	Apparently, some people feel that running out of free IPv4 addresses
>> 	will be bad.  This proposal provides a self-regulating mechanism
>> 	for delaying that date by gradually increasing the time it takes to
>> 	receive a new allocation or assignment from ARIN as the free
>> 	pool shrinks.
>> 	The fewer free addresses there are, the longer it will take to
>> 	receive an assignment or allocation.  Further, it has the additional
>> 	advantage that it will provide faster allocations and assignments
>> 	to those organizations which have smaller existing IP resource
>> 	pools from which to draw while extending the delay factor for
>> 	those few organizations which are already consuming the vast
>> 	majority of IP address resources.

		Defining X -- This policy is deliberately incomplete.  It is
		expected that the AC will judge consensus of the community
		towards a value of X and, at the time of promoting this policy
		to last call will replace the text "X shall be defined as _____.
		(See rationale section)." with "X shall be defined as <number>"
		where <number> is replaced with the actual number judged
		to be the consensus of the community.

		The policy author believes that 400 is a good starting value
		for X.  The community is encouraged, when commenting on
		this proposal to state their desired value for X as part of their
		comments.  Higher values of X will create a steeper ramp
		and longer initial delays.  Lower values of X will create a
		more dramatic curve that starts to pick up later.

>> 9.	Timetable for implementation: April 1, 2008
>> 10.	Meeting presenter:  Ruben Garret L.  Goldberg
>> _______________________________________________
>> You are receiving this message because you are subscribed to the  
>> ARIN Public Policy
>> Mailing List (PPML at arin.net).
>> Unsubscribe or manage your mailing list subscription at:
>> http://lists.arin.net/mailman/listinfo/ppml Please contact the  
>> ARIN Member Services
>> Help Desk at info at arin.net if you experience any issues.

More information about the ARIN-PPML mailing list