Probably because of blocking at the origin point, such as corporate net-mgrs trying to prevent bandwidth hogs or liability issues.
Sure but my point is, that unless you run your private p2p network somewhere which is not connected to the internet, you´ll end up with similar figures because these "net-mgrs" will be out there doing their thing and there is nothing you can do about them doing it. Pete
Rubens
----- Original Message ----- From: "Petri Helenius" <pete@he.iki.fi> To: "Stephen Sprunk" <stephen@sprunk.org>; "Jack Bates" <jbates@brightok.net> Cc: "Richard A Steenbergen" <ras@e-gerbil.net>; "Peter Galbavy" <peter.galbavy@knowtion.net>; "Mike Lyon" <mlyon@fitzharris.com>; "Simon Lyall" <simon.lyall@ihug.co.nz>; "Tony Rall" <trall@almaden.ibm.com>; "North American Noise and Off-topic Gripes" <nanog@merit.edu> Sent: Monday, March 31, 2003 6:08 PM Subject: Re: State Super-DMCA Too True
| | > Well, most p2p apps live on well-known ports, and Cisco's QOS mechanism | > allows easy classification on ports. Yes, most of the p2p apps are | > port-agile -- but only if they are completely blocked. My experience is | > that if you let the p2p stuff through, it'll stick to its default port and | > you can police with impunity. | | Our data shows that between 30% and 50% of p2p data flows on "non-standard" | ports if you run an unblocked environment. | | Pete |