Re: Traffic ratio of an ISP

[Prasun Dey <prasun () nevada unr edu>]

Hi Job,
While doing some study, I recently came across this 
https://drpeering.net/white-papers/The-Folly-Of-Peering-Ratios.html
This discussion was from from a Nanog meeting that took place a long time ago. This made me interested to know whether 
there is some actual numbers behind those PeeringDB traffic ratio labels. 
I think your comment on the importance of traffic ratio for a specific ASN pairing is spot on. Those information are 
confidential, and rightly to be so. All I wanted to know how much traffic a provider handles (receives vs. delivers), 
regardless of its business type. As other members have also mentioned, general consensus is, CPs are outbound, while 
transits are Balanced. I was wondering if there is some publicly available information about this labels. But, seems 
like these are more like generic information and their impact is very small in real life while ISPs decide to peer.
Thank you for your response.

-
Prasun

Regards,
Prasun Kanti Dey
Ph.D. Candidate,
Dept of Electrical and Computer Engineering,
University of Central Florida
web: https://prasunkantidey.github.io/portfolio/

> On Jun 20, 2019, at 10:27 AM, Job Snijders <job () instituut net> wrote:
>
> On Thu, Jun 20, 2019 at 4:21 PM Steller, Anthony J
> <Anthony.Steller () charter com> wrote:
> > because it really don’t matter in the whole scheme of things.
>
> Indeed, it doesn't matter. The "traffic ratio" field in PeeringDB
> probably should be deprecated, there is no formal definition nor is
> are there any operational consequences to changing the contents of
> that field. The contents of the field are entirely arbitrary.
>
> If the traffic ratio is relevant (I am not saying it is or isn't),
> such traffic ratios probably should be viewed in exclusively in
> context of specific ASN pairings. Maybe between you and me we'll see
> the dominant traffic direction being one way, and with another ASN
> pairing we see the opposite. There is no telling other than through
> observation, any such observations are unlikely to be shared with the
> general public.
>
> Kind regards,
>
> Job