[arin-ppml] ARIN Multiple Discrete Networks Policy

John Curran jcurran at arin.net
Sat Oct 1 18:45:32 EDT 2011


On Oct 1, 2011, at 6:15 PM, Richard A Steenbergen wrote:

> On Sat, Oct 01, 2011 at 09:06:58PM +0000, John Curran wrote:
>> 
>> Looking at the proposed definition:
>> 
>>   Multiple = more than one network.   
>> 
>> This is fine; no further elaboration needed.
>> 
>>   Discrete = "one which must implement a unique routing policy"
>>               (distinct from the other networks)
>> 
>> How does ARIN staff determine that the network "must implement a 
>> unique routing policy"?  
> 
> If they meet one of the existing examples for compelling reasons to use 
> multiple discrete networks? I kinda figured that part was a given. :)

Richard - Hopefully, we don't need examples of compelling reasons if
a clear technical definition of "discrete networks" is added (as it 
would remove the requirement for judgement of "compelling reason")

>> As written, I see that this can be read as either [A] "the requestor 
>> must clearly show that the network is configured with routing policy 
>> which is unique from the others" or [B] "the requestor must show that 
>> the network cannot function correctly without a routing policy which 
>> is unique from the others" Which of the above ([A], [B]) did you 
>> intend?
> 
> I'd say both, but again I think the policy already says this. :)

[A] is a simple matter of fact, i.e. show networks with unique routing
policies.

[B] still requires judgement, i.e. explain why the networks must have 
unique routing policies in order for the networks to operate "correctly".  
It is clearer than present policy since it would have a clear definition
that equates discrete networks as networks with unique routing policies, 
but wouldn't eliminate the need for judging whether or not the use of 
those unique routing polices was truly necessary.

Using [A], if a definition of "Multiple Discrete Networks" is added to 
read: "Multiple networks are networks with unique routing policies",
then the criteria could be greatly simplified to simply show presence
of multiple discrete networks.

Note: I have no idea whether the community wants either change, 
but suggest you propose either one, should it better meet your 
needs.    Details on submitting a policy proposal are here: 
https://www.arin.net/policy/pdp_appendix_b.html

Thanks!
/John

John Curran
President and CEO
ARIN




More information about the ARIN-PPML mailing list