NoCashBob Posted May 15, 2008 Report Share Posted May 15, 2008 I've seen a few posts about that claim some ISPs may be using as a criteria for throttling the many connections to a given port torrents use. I don't know whether or not this is true, but it seems very plausible that they may use this in combination with header inspection as well as other methods unknown to me.What I wonder is whether or not a BitTorrent client could work around it by incorporating a (very complex to code I'm sure) option to have the program use multiple ports. Having an option for the user to designate how many connections per port they want to allow could be good too (to allow them to adjust it to the maximum value before causing trouble).Now This would depend on having a header encryption capable of fooling the ISPs DPI, would depend on said ISP actually telling the truth on whether their DPI solution really does only inspect the hear (doubtful) and finally that there are no other measures in place. Link to comment Share on other sites More sharing options...
Firon Posted May 15, 2008 Report Share Posted May 15, 2008 No ISP goes port-based anymore. Link to comment Share on other sites More sharing options...
NoCashBob Posted May 15, 2008 Author Report Share Posted May 15, 2008 Not even as a secondary measure? Link to comment Share on other sites More sharing options...
Switeck Posted May 15, 2008 Report Share Posted May 15, 2008 Maybe the occasional 3rd world country low-end broadband "ISP" will still do that, but they're about the only ones. Link to comment Share on other sites More sharing options...
hermanm Posted May 15, 2008 Report Share Posted May 15, 2008 In Bittornado (the version TorrentFlux can run), you specify a range of ports in your configuration. Each torrent receives a unique port number. Link to comment Share on other sites More sharing options...
Ultima Posted May 16, 2008 Report Share Posted May 16, 2008 Uh huh...? Link to comment Share on other sites More sharing options...
jewelisheaven Posted May 16, 2008 Report Share Posted May 16, 2008 ... one-port-per-torrent was in back in 03 or 04? It's been way too long to be in memory for me. Why would one choose to spread that load back out? In the context of trackers it makes sense to be able to rotate ports for announces, as trying to theoretically keep track of hundreds of thousands of hashes and associated IP:PORT combinations could lead to bloat and therefore slowness. It's still a good idea to NOT use default ports in-case you get stuck with some archaic filtering, but you really are more likely to be throttled/disconnected based upon the activity whether or not it's on some well-known port such as HTTP (if ISPs EULA block you from running an http server...) or something above the reserved 1024. Link to comment Share on other sites More sharing options...
Lord Alderaan Posted May 16, 2008 Report Share Posted May 16, 2008 Back in those days one instance of the listener was run for each torrent (usually because a whole instance of the client was started for each torrent). Advancing to a single listener for all torrents was simply an improvement. There is no good reason to revert back to multi-port setups nowadays. Link to comment Share on other sites More sharing options...
Switeck Posted May 16, 2008 Report Share Posted May 16, 2008 It is often forgotten that even IF your listening port is just one port...you're typically using a whole horde of outgoing ports. Typically, these are ephemeral ports (about 1000-5000 in range), but uTorrent can be configured to use almost any consecutive port range you tell it to now.If you are firewalled, then ONLY your outgoing ports are really being used anyway...though your ISP may see incoming packets you never get. Link to comment Share on other sites More sharing options...
Lord Alderaan Posted May 16, 2008 Report Share Posted May 16, 2008 True. But in the past ISPs throttled based on the destination port. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.