[arin-ppml] ARIN-PPML Digest, Vol 106, Issue 8

[Owen DeLong]

On Apr 4, 2014, at 11:00 AM, Steven Ryerse <SRyerse at eclipse-networks.com> wrote:

> If an org with no resources applies they should at least be able to get the minimum which has been set by this community which I think is currently at a /22.  Always!

Depends… End-user /24, ISP multi-homed /22, ISP non-multi-homed /20 IIRC.

> If an org wants larger than a /22 they need to be able to demonstrate in a reasonable way that they are a larger org with a network size that justifies a larger allocation.  The first way is what allocation do they already have?  If they have say a /19  or equivalent maybe they can demonstrate they need say another /19 by furnishing to ARIN maybe their financials and the investment they have actually made to justify another /19 or whatever.  (I'm just using the /19 as an example.)

Why would financials or investment have any direct correlation with network size in any organization, let alone all organizations?

> Organizations supply financials to Banks all the time and this and supplying network info can be a similar process.  If an org can prove they just spent 50 million on a new data center then that should be justification to get an allocation of a size to run that size data center. 

How do you judge the size of a datacenter in IP resource requirements based on the amount of money spent on the datacenter? A relatively small datacenter with a lot of virtual hosts on a small number of machines might need a whole lot more IP addressing than a vastly more expensive datacenter housing a small number of supercomputers and their ancillary systems, costing 100s or even 1000s of times more money.

> So if we want the current minimum we should just apply and it should be allocated, and as long as we keep paying our fees to ARIN it is ours to use.  If we stop paying that block goes back into the pool that ARIN has to allocate.  

So if I stand up a bunch of shell companies, they should each be able to get a /22 and I should be able to repeat that exercise until I run out of money or desire for address space? An interesting theory.

> If we want a /19 or a /16 then we need to satisfy ARIN that we are an org of that size and have a network that justifies a /19 or a /16.  And maybe prove the actual expenditures. 
> 
> And if we want a /9, regardless of what we have now, we're gonna really have to provide solid info that justifies the size of our org and the size of our network justifies a /9.  

These two statements do not seem to differ from current practice, so I’m not sure what sort of change, if any, you are advocating for in this case.

> This is not rocket science. It would however require the input of the many knowledgeable members of this community to help determine what is require at each level.  I think this community could handle that just fine!

I think we have… How do you think the existing NRPM 4 and NRPM 6 (and the rest of NRPM for that matter) have come about?

Owen