Re: Geoff Huston on IP address space utilisation in 2015.

Hi All,

With regards to IPv4 running out, and with regards to my previous statements about AfriNIC.

I felt its worth making this list aware of the following policy which was just posted on the AfriNIC policy list.  The draft of the new “soft landing” policy is pasted below, and I think it makes clear that in some of our views, IPv4 depletion is coming, and its time to let it come.

Just for some background before the proposed policy is pasted below.

Under current AfriNIC policy, when AfriNIC is down to the last /8 (which is likely to happen by July/August at current allocation rates), the allocation rules start to change, and getting space becomes more difficult and in smaller blocks, and when it gets to the last /10 it gets even tighter.  The original policy that put these rules in place was designed to effectively keep IPv4 alive for a far longer time.

Myself and the other policy authors feel this will hurt Africa, so we are proposing a repeal of that policy with a replacement.

I HIGHLY suggest that every ISP and every person on this list who has, or plans to get IPv4 space in future read this stuff carefully, it could have a MAJOR impact on your businesses going forward.  I’d also HIGHLY suggest that you all get on the RPD list and have your say on the policy discussions going on there, and better yet be at the next Afrinic meeting in Gaborone in Late May, because the decisions taken there WILL impact you, and you have the right to have a say in your own futures.

The new draft is as follows (and should be read in conjunction with http://www.afrinic.net/en/library/policies/697-ipv4-soft-landing-policy)

Thanks


Andrew

—— Policy Draft follows —


AFPUB-2016-V4-002-DRAFT01
Soft Landing Overhaul
  
  Authors:
  a.    
  Andrew Alston, [email protected]
  b.    
  Kris Seeburn, [email protected]
  c.    
  Mark Elkins, [email protected]
  d.    
  Michele McCann, [email protected]
  e.    
  John Walubengo, [email protected]
   
  Draft Policy Version:       01
  Submission Date:            12 February 2016
  Status:                     Under Discussion
  Amends:                     AFPUB-2010-v4-005 (IPv4 soft landing policy)
  
  
1.      Summary of the problem being addressed by this proposal
At the time when the original soft landing policy was authored, there were many unknowns and 
circumstances that could not be foreseen, and as a result of this, the policy in its current 
form may actually damage the interests of the AFRINIC community rather than assist it.
Primary among these, it was not known when the rest of the world would run out of IPv4 space, 
and the adoption rates of IPv6 were also an unknown quantity.
While it is acknowledged that there is a need to ensure that new entrants into the IP world 
may require some small amount of IPv4 space, beyond this, further delaying the depletion of 
IPv4 address space may well be holding the region back while the rest of the world moves on, 
leaving Africa at a significant disadvantage moving forward.
In the original policy replaced by this, the numbers and allocation levels were also not based 
on any fundamental justifications, because of the unknowns that existed at the time.

2.     Summary of how this proposal addresses the problem
This proposal still maintains a block of space reserved for new entrants, but beyond that, it 
allows for the natural depletion of IPv4 through standard demand, and hence encourages the uptake 
of IPv6 in a more aggressive manner.

3.     Proposal

3.1.  Section 3.2 is amended to remove the definition of “Final /8 block of IPv4 address space, or “Final /8”

3.2.  Section 3.2 is further amended to add the following definition: 
“New Entrant Block - A /13 block of IPv4 space, reserved in entirety, for allocations of space to 
members of AFRINIC that at the time of application have no previous IPv4 address allocations.   A /13 
was chosen based on historical member growth numbers by AFRINIC including a certain increase in those 
allocations to provide sufficient space to allocate to new members of a period of 2 years”

3.3.  Section 3.2 is further amended to remove the definition of “New LIR”

3.4.  Section 3.2 is amended to add the following definition: “Additional and reclaimed space: 
All IPv4 address blocks recovered from non paying clients, as well as all allocations of address space 
made to AFRINIC by IANA or a replacement organisation”

3.5.  Section 3.2 is further amended to add the following definition: 
“New Entrant” – Either a member or new member that at the time of application had no previous IPv4 
allocations or assignments made to them by AFRINIC, and were not holders of legacy IPv4 space or 
other IPv4 space sourced either through a potential transfer market or other RIR.

3.6.  Section 3.4 is amended to read:
The “Current Phase” is the status-quo at the time of the adoption of this policy.  During this phase, 
AFRINIC will continue allocating or assigning IPv4 address space to LIRs and End Users using current 
IPv4 allocation policies as determined by the community through the policy development working group.
The current phase will continue until the depletion of IPv4 address space occurs, with the exception 
of IPv4 reservations as defined by this and other currently in force policies.

3.7.  Section 3.5 is renamed to “New Entrant specification”

3.8.  Section 3.5 is amended to read:
At the time where an application is made that will not be fulfilled out of the AFRINIC pool, with the 
exclusion of space reserved by this and other policies, the only applications for IPv4 space which will 
be further considered by AFRINIC will be for New Entrants. The maximum size of a New Entrant allocation 
will be a /22. New Entrant applications will be processed on a first in first out (FIFO) basis, that is 
To say that applications will be processed in the order in which they are received.
New Entrant applications with regards to justifications must conform to current IPv4 allocation policies 
as defined by the community.

3.9.  Section 3.5.1 is amended to read:
All space falling under the definition of Additional and reclaimed space, as from the time of ratification 
of this policy, will become part of the new entrant vlock and will be reserved for members who meet the 
New Entrant definition.

3.10.  Section 3.5.2, 3.6, 3.7, 3.8 and 3.9 are revoked.





On 12/02/2016, 11:09 PM, "[email protected] on behalf of Hendrik Visage" <[email protected] on behalf of [email protected]> wrote:

>On Fri, Feb 12, 2016 at 6:27 PM, Joe Abley <[email protected]> wrote:
>>
>> On 1 Feb 2016, at 03:29, Nishal Goburdhan <[email protected]> wrote:
>>
>>> v4 run out is inevitable.  sorry, but if you’re going to waste your time trying to fix this in policy, in an environment with *very* low SNR, you’re just trying to save a very leaky bucket.
>>
>> Long-term viability of the network is one thing. Where the cash associated with the single remaining supply of a scarce resource goes is another; it can either leave the continent or it can stay.
>>
>> This reminds me of a similar tension in Canada relating to oil sands development: on the one hand there's an inevitable future that is not driven by carbon, so why waste time developing what is a very resource-intensive way of pulling it out of the sand; on the other hand, as the oil runs out and demand peaks over supply, there's money to be made that makes the investment arguably more worthwhile than it has ever been before.
>>
>> In the case of oil there's an ecological argument for leaving the carbon in the ground. When it comes to v4 markets, there are no boats of bearded activists defending the pristine sanctity of 32-bit integers.
>>
>> However, there are also established national and regional regulators that control the flow of money resulting from the oil extraction -- paraphrasing Andrew's call to arms, perhaps these particular integers are continental-scale national resources, and need some regulation that reflects that.
>
>The problem with those "regulations" and "money flows", is that it
>just prolongs the inevitable... and by implication, our suffering and
>hurt... if we have IPv4 runs out in totality, then it'll be a way to
>start getting our consumers requesting IPv6 from the 3G, ADSL and
>fibre providers. Then we'll have a case where Telkom/etc. are either
>forced to getting IPv6, or we'll have to handle the pains of trying to
>preserve IPv4 for a tad longer, and then people starts to move
>elsewhere where they do get IPv6 etc.
>
>I'm totally in favor of letting the IPv4 run out and having people pay
>through their ears together with an arm and a leg for it... thereby
>gently forcing then to go the "costs savings" route to get IPv6
>equipment installed and running
>
>We in Africa are just a bit blessed to actually have an abundance of
>available IPv4 given our lagging behind the tech curves (or rather us
>being more consumers than producers of content). That same abundance
>is our curse, as that'll not force our ISPs neither our national ADSL
>carrier nor our cellular providers to start deploying and testing IPv6
>for the masses...
>
>/me also want to buy a /8 to do his part in the depletion
>acceleration... just need the monies to buy that for a start ;(
>
>_______________________________________________
>IOZ mailing list
>[email protected]
>http://lists.internet.org.za/mailman/listinfo/ioz
_______________________________________________
IOZ mailing list
[email protected]
http://lists.internet.org.za/mailman/listinfo/ioz