NNSquad - Network Neutrality Squad
[ NNSquad ] Re: pcap files of the Comcast forgeries?
John: For BitTorrent http://torrentfreak.com/images/comcast-rst1.txt Fot ED2K http://www.dslreports.com/forum/r18926539-Comcast-P2P-Mgmt-Wireshark-eMule-Example Lauren: I can explain the conflicts. My original post in May http://www.dslreports.com/forum/remark,18323368 was incorrect on that fact. Originally, I thought it was Comcast-to-Internet, first because the Sandvine documentation I read during my research suggested that the installation would occur at the network boundary; and second because I had not ran across a Comcast-to-Comcast situation that was aborted (TCP RST). It was only later that I ran across two instances of Comcast-to-Comcast interference and decided I needed to change my original statement: http://www.dslreports.com/forum/r18918622-Re-Comcast-is-using-Sandvine-to-manage-P2P-Connections -- which is explained in http://www.dslreports.com/forum/r18936691-x that the device isn't at the network boundary in Comcast's case, but it located in the metro areas. Ironically this means that Comcast-to-Comcast P2P connections (those which should be the most affordable and the least likely to cause congestion) are twice as likely to be aborted. --Robb On 12/17/07, John Bartas <jbartas@clear2net.com> wrote: > Does anyone have some pcap files of the Comcast/Sandvine forgeries? It > might help us on the Working Group to think of better ways to detect > this sort of activity. > > -JB- > > [ I might add, is there general confirmation that Comcast's P2P > reset spoofing behavior was limited to Comcast <-> Comcast > situations, or were Comcast <-> Outside connections also involved? > I'm hearing conflicting info on this. Thanks. > > -- Lauren Weinstein > NNSquad Moderator ] > -- -- Robb Topolski (robb@funchords.com) Hillsboro, Oregon USA http://www.funchords.com/