[arin-ppml] Draft Policy 2009-1: Is there an Emergency?

Tue Mar 31 17:17:16 EDT 2009

On Mon, Mar 30, 2009 at 00:05, David Farmer <farmer at umn.edu> wrote:
> I would like to motivate a discussion of the question "Is there
> an Emergency?"
>
> I have heard several people express the opinion that they don't
> see an emergency.  I would like to respectfully disagree with
> that opinion.
>
> In my opinion the crux of the emergency is IPv4 exhaustion
> combined with the lack of IPv6 adoption, which means we are
> going to hit the proverbial wall when it comes to functional IP
> address availability.  But when does this become an
> emergency?
>
> Maybe we can use a car accident as a metaphor;  When does
> a car accident start?  When you hit the wall?  When the
> airbags deploy?  When you fail to make the turn or hit the
> breaks in time to prevent yourself from hitting the wall?
>
> Using this metaphor, I propose;  IANA free pool exhaustion is
> equivalent to the car hitting the wall.  The trigger set in 2009-2:
> Depleted IPv4 Reserves, is the equivalent of the airbags going
> off shortly after the car hits the wall.  RIR and ISP free pool
> exhaustion are equivalent to the passengers hitting the interior
> of the car and the brain and internal organs colliding with the
> skull and the rest of the body, receptively.
>
> So when did the IPv4 car accident start, when did we hit the
> point where we would inextricably hit the wall?  I'm not exactly
> sure, but I think most of us started realizing back in 2007 that
> we were going to inextricably hit the wall.  And today, to me
> personally it is virtually unquestionable that will we are going to
> hit the wall.  We obviously haven't hit the wall just yet, but the
> car is headed toward the wall to fast to stop or turn, the
> accident must and will happen.

I am not convinced that the car hitting the wall analogy is of much
use when discussing IPv4 free-pool exhaustion, mainly because I can't
seem to envision many organizations (if any) exploding at IANA, RIR,
or ISP free IPv4 runout.  Perhaps a car running out of gas is a more
applicable comparison?  I know it lacks the drama of a giant fireball
but it seems more accurate.  Organizations unable to obtain more
address space will not cease to exist, they will simply cease to grow
(unless they adopt an alternative source of IP / fuel).  Also, I have
a very hard time determining the location of a single collective car
along this path - as someone (please forgive me for not looking up
who) astutely pointed out in a previous ppml thread; IPv4 free pool
exhaustion has effectively already happened for many orgs who will not
be able to justify an allocation in the near future (2- years) but may
need more addresses in the longer term.

So there are a bunch of vehicles all moving along fueled (almost
exclusively) by IPv4 and if nothing is done, they will stop moving.
So the emergency is not a collision but a stall and when that happens
is largely dependent on each individual vehicle's actions.  Mega ISP
is akin to a bus carrying many customers along with them and thus
burning fuel quite quickly; but they have a lot of it.  Mom's Shop is
more like a hybrid; carrying little fuel but only sipping it along the
way.  And ARIN is pit lane, the gas station -- with one obvious
difference; IP is not truly consumed so those no longer needing it to
fuel their car can return it to the station (or transfer it to
others).

>From this perspective the emergency begins for each vehicle when they
reach a point where they _will_ stall.  I think the better question
though (and the one you are asking) is when does it become an
emergency for the community as a whole?  When is it an emergency for
the ARIN region?  Unfortunately, to this I do not have a definitive
answer.  Is it when we know that any one member will stall?  Or when
10% reach that point?  25%, 50%?  What portion of our community is at
this point already?

The thing I wonder about most though, is when the BoT declared this an
emergency, were they speaking about the ARIN community or ARIN itself
(or maybe both)?

~Chris

>
> Further, it is possible we don't have as much time as we think
> we do.  We currently have approximately 500 Million IPv4
> address in the IANA Free Pool.  While current projections,
> based on current usage rates, provide a little over two years to
> exhaustion[1].  However, it is not difficult to imagine scenarios
> where the IANA free pool could be exhausted much sooner
> than that.  For example, if mobile providers were to start
> issuing IPv4 address to mobile hand sets it wouldn't be hard to
> exhaust the IANA free pool in no time flat[2].
>
> [1] http://www.potaroo.net/tools/ipv4/index.html
>
> [2]
> http://newsroom.parksassociates.com/article_display.cfm?articl
> e_id=5128
>
> I'm not saying that will or even should happen, but it is by no
> means impossible.  Further, under current policies if the mobile
> industry came to the RIRs for IPv4 addresses for hand set, the
> RIRs would likely have to fulfill the requests, and exhaust the
> IANA free pool in short order.
>
> Therefore, at least in reference to IPv4, I believe there is a
> valid Emergency.
>
> So, I'm interested to hear other people's opinion on if there is
> an emergency.
>
> ================================================
> =======
> David Farmer                                 Email:
> farmer at umn.edu
> Office of Information Technology
> Networking & Telecomunication Services
> University of Minnesota                      Phone:     612-626-
> 0815
> 2218 University Ave SE                       Cell:
> 612-812-9952
> Minneapolis, MN 55414-3029                   FAX:       612-626-
> 1818
> ================================================
> =======
>
> _______________________________________________
> PPML
> You are receiving this message because you are subscribed to
> the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
> Unsubscribe or manage your mailing list subscription at:
> http://lists.arin.net/mailman/listinfo/arin-ppml
> Please contact info at arin.net if you experience any issues.
>
-- 
Chris Grundemann
weblog.chrisgrundemann.com