[arin-ppml] Set aside policy?

Jason Schiller schiller at uu.net
Fri Jul 2 10:49:18 EDT 2010


In the very near future people will be forced into building IPv6-only 
networks.  The question is how will these IPv6-only networks reach the 
content that is only available on the IPv4 Internet.

Personally I don't think the IPv4/IPv6 Carrier Grade NAT solution is going 
to work well.  I suspect it will have problems with scaling, problems with 
being easily supported, problems with providing good CALEA support, 
interfere with geo-location, and be a big target (with lots of collateral 
damage) for DoS attacks.

The next less bad solution is to offer exactly one IPv4 address to each 
IPv6-only network.  This will allow them to stand-up their own (read at 
the customer premise) IPv4/IPv6 NAT.  This does not have the same scaling, 
CALEA, and support issues, and has the other problems to a lesser extent.  
Hopefully, these addresses are part of an aggregate and people are not 
attempting to get IPv4 PI /32s routed on the Internet.

While this level of rationing will save lots of addresses from business 
customers, it will be business as usual for most consumer customers.  I'm 
not sure what to think of that.

There are some other questions here.  If a business customer has multiple 
connections and a need to do traffic engineering, could they get one IPv4 
address per connection?  What if they are IPv6 multi-homed, will they need 
one globally unique IPv4 address for Router ID on each eBGP speaking 
router?  What if for reasons of scale they need to offload the NAT 
function to a purpose-build box sitting directly behind their router, in 
that case can they get enough IPv4 addresses for their router and their 
NAT box?

The bigger question is what do you consider IPv6 critical mass, for 
example when 80% of Internet facing services support IPv6?  And how long 
until that occurs, for example 5 years?  (pick whatever numbers make sense 
to the community)  In that amount of time, how many new IPv6-only networks 
will you activate?  

Unfortunately this line of logic may lead one to conclude that the ARIN 
community needs more than a single /8.  In that case is it wise to have 
an ARIN policy that allows people to get "Transition Space" for this 
purpose only now, before we whittle the IANA Free Pool down the last 5 /8s 
that are under the "Global Policy for the Allocation of the Remaining 
IPv4 Address Space"?  Note that this will greatly accelerate the depletion 


Jason Schiller                                               (703)886.6648
Senior Internet Network Engineer                         fax:(703)886.0512
Public IP Global Network Engineering                       schiller at uu.net
UUNET / Verizon                         jason.schiller at verizonbusiness.com

The good news about having an email address that is twice as long is that
it increases traffic on the Internet.

On Fri, 2 Jul 2010, Martin Hannigan wrote:

|Date: Fri, 02 Jul 2010 10:11:47 -0400
|From: Martin Hannigan <marty at akamai.com>, marty at akamai.com
|To: Scott Leibrand <scottleibrand at gmail.com>, arin-ppml at arin.net
|Subject: Re: [arin-ppml] Set aside policy?
|On 7/1/10 8:48 PM, "Scott Leibrand" <scottleibrand at gmail.com> wrote:
|> On Thu 7/1/2010 5:45 PM, marty at akamai.com wrote:
|>> Is there _anything_ that people would consider reasonable for a set-aside
|>> policy? I'm thinking about policy for transition infrastructure and wanted
|>> to see if anyone else is thinking about this.
|> How much space are you thinking about reserving?
|That's a premature question IMHO. The concept of what transition
|infrastructure "TI" is needs to be hammered out. Probably the hardest part
|of any set-aside policy that will make an impact.
|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:
|Please contact info at arin.net if you experience any issues.

More information about the ARIN-PPML mailing list