[arin-ppml] IPv6 Multihomed networks

[David Farmer]

Kevin Kargel wrote:
>> On 2010.02.22 19:59, Owen DeLong wrote:
>>> I cannot speak for the entire AC, however:
>>> I'm all for relaxing the policies for routable space. I'm opposed
>>> to replacing them with a policy which was put in place under the
>>> guise of creating "non-routable" space.
>>>
>>> Making assignments for various things out of "a block reserved for
>>> that purpose" essentially creates an artificial class system.
>> Then there should be no distinction between 'private' and 'public', and
>> there should be no difference in policy requirements for acquiring either.
>>
>> imho, if this is the case, then the entire discussion of "Non-connected
>> networks" is futile, isn't it?
>>
>> I mean, if you already know that the path of least resistance is already
>> going to be allowed, why bother creating policy?
>>
>> Steve
>> _______________________________________________
> 
> This is my point exactly in all of this.  To my way of thinking you should be able to apply for and receive a block of IPv6 because you have some need for it.  

I agree, but I believe it is not unreasonable for you to explain that 
need to some extent to ARIN.  I don't want to make it a burden, but some 
checks and balances are usually a good thing.

> Think of that block as containing both routed and non-routed space and 'wow', you get to decide for yourself how much you want to route globally and how much you don't want to route, then you adjust your advertisements and firewalls accordingly.  
>
> You get the space, you can route it or not route it, you set the borders, and you can change the borders whenever you want. 
>
> So long as you keep your contact information current and pay your fees your assignment remains valid, whether you actually route it or not.
> 
> I see no reason for setting aside special non-routed space in the pool when you can already use your assigned space and you don't have to route it.
> 
> If we did create a separate pool for some reason, then IMHO it depletes the pool the same as any other reason and should have the same requirements.

I mostly agree with this, and if I put on my "I believe in RPKI" hat on 
then no problem.  But, if RPKI doesn't happen, then this non-routed 
space could create abuse problems, and make a place for bad guys to make 
a home. So I think this has been an important discussion.

> So yes, as you said "the entire discussion of "Non-connected networks" is futile"

I'll just say that I don't think this discussion has been futile at all, 
  I believe it may be heading us toward a consensus.  And, if we finally 
get to a consensus then I believe it has been a very productive discussion.

I have no illusions that PP#101, PP#106 or PP#107 will end up as policy 
as they are currently written.  Sometimes the important thing is to put 
things in front of people and let them take their shots at them.

So please keep shooting at them.

-- 
===============================================
David Farmer               Email:farmer at umn.edu
Networking & Telecommunication Services
Office of Information Technology
University of Minnesota	
2218 University Ave SE	    Phone: 612-626-0815
Minneapolis, MN 55414-3029   Cell: 612-812-9952
===============================================