[arin-ppml] ARIN Multiple Discrete Networks Policy
Richard A Steenbergen
ras at e-gerbil.net
Sat Oct 1 01:11:04 EDT 2011
On Sat, Oct 01, 2011 at 03:15:20AM +0000, John Curran wrote:
> I did not say that it has no effect on aggregation. I said that the
> point of the policy is not to improve aggregation; the policy exists
> to allow organizations to readily obtain address space as a single
> organization under criteria applicable when there are multiple
> discrete networks for a compelling reason.
> I'd recommend either redefining "multiple discrete networks" to have a
> clear technical definition in the policy, or refining the existing
> examples so one of them directly meets your specific needs (e.g.
> removal of the term "autonomous" or changing it to an AS reference
Ok, let me see if I can summarize this.
You're saying that if someone has a compelling reason for running
multiple discrete networks, it means that they are incapable of
announcing their single allocation as an aggregate, and therefore need
the MDN policy to provide multiple blocks which they CAN announce as
You then go on to say that if someone benefits from the aggregation
being provided in the previous statement, since aggregation is not an
intended effect of the policy they therefore don't qualify to use it.
Do you not see the circular logic here? I feel like I need to flowchart
this or something. :)
Let's think about this:
* There is already an existing policy which enumates a very specific
list of example "compelling reasons" for running multiple discrete
* One assumes that if you can exactly match an example from the list of
"compelling reason" included in there policy, you therefore do infact
have a compelling reason for running them.
* Thus, your compelling reason is THE EXAMPLE THAT YOU MATCHED FROM THE
LIST OF COMPELLING REASONS, not the aggregation benefits that may
come as a result.
Is there something else that I'm missing? I could at least understand
and correct it if there was something specific like "I don't think you
meet compelling reason X", but I'm having a really hard time arguing the
simple chain of logic above any clearer than I already have.
Richard A Steenbergen <ras at e-gerbil.net> http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)
More information about the ARIN-PPML