I have been probing this issue of BitTorrent anonymity and I have made some conclusions but not final conclusions.

I'm aware that PPTP VPNs have security flaws for a variety of reasons; and that, in regard to P2P sharing, PPTP is of no use because of IPv6.

I'm also aware that configuring the network proxy connection of BitTorrent is of no help in anonymizing P2P. Like VoIP connections, P2P seem to ignore or bypass network proxy settings. Part of this may have something to do with IPv6, part of it with UDP, and part of it the protocol stack.

However, like the issue with Skype and other VoIP services, P2P seems like it can be anonymized, via JonDonym and OpenVPN, though no absolute conclusion can be rendered.

Can JonDonym presently handle IPv6? I know Proxifier and ProxyCap have full support for IPv6 and SOCKS5 can handle UDP...when playing with P2P BitTorrent it seems that the data is anonymized.

I will have to do a more in depth network protocol analyzer.

I will do more research on this matter.

P2P can not be anonymized with onion encryption services like Tor or JonDonym. Have a look at the paper "privacy attacks for bittorrent over tor" http://hal.inria.fr/docs/00/47/15/56/PDF/TorBT.pdf

May be VoIP can be anonymized via JonDonym in furure, we need some reearch. It seems, a VoIP proxy behind the last mix is requiered. May be, JonDonym mix operators can offer a VoIP proxy or a public VoIP proxy can be used?

No, but we hope IPv6 support is ready in ... - hmmm - soon.

Thank you Cane for the paper. However, I think the paper is wrong.

I think there are only IP leaks or exposure if you merely change the network connections...

Here is what I have been able to document:

1. BitTorrent is tunneled (rerouted) by a socksifier through Tor or JonDonym mix servers.

2. SOCKS5 supports UDP, though how well is unknown.

3. Proxifier currently does not support UDP. Therefore, when you tunnel BitTorrent data packets through Tor or JonDonym mix servers, UDP appears not be forwarded.

4. ProxyCap does support UDP via SOCKS5, though how well is unknown. If SOCKS5 supports UDP, then ProxyCap should forward UDP and TCP data packets through JonDonym mix servers, assuming JonDonym can support UDP.

5. I have done BitTorrent IP tests on a Tor VPN and have confirmed the following:

UDP was not tunneled (probably because Proxifier does not support UDP).

The BitTorrent IP was in fact anonymized (Tor IPs). My real IP never leaked out; not during any part of the test, even when I was downloading the featured video.

See the test yourself:

http://www.checkmytorrentip.com

However, the test does not work with JonDonym--I do not know why.

I tested an OpenVPN and it also anonymized BitTorrent--but the test indicated that UDP was forwarded.

I would surmise that if you or anyone else performed a BitTorrent IP test with ProxyCap, enabling UDP and TCP, the test would probably also indicate UDP.



This really does not make any sense to me. I do not think this is accurate. Why would an extra mix server be needed? If the data traffic is tunneled via a socksifier, the data traffic should be anonymized.

Everything seems to point to this...



In the mean time, the new edition of Proxifier can tunnel IP4 connections through IPv6 and IPv6 through IPv4.

You can create a forward connection through the mix cascades but you are not reachable behind the mixes. For connections from other people to you you need a SIP proxy behind the exit mix, not an extra mix server.


BitTorrent with JonDonym is very expensive. We think, it is not interesting for JonDonym users. You have to pay for down- and upload. Only download your stuff and afterwards close the connection is not the fair use of BitTorrent.

If the paper was wrong please contact the writers. JonDonym user are affected in the same way like Tor users by this privacy attacks.

The website http://www.checkmytorrentip.com does not check the issues from the paper. It checks only the connected IP address.

Got it. But I am still struggling with this.



First, I neither like nor use BitTorrent. I was only experimenting with it, to see what kind of results would come. Yes, it would be expensive but even if it wasn't it still would not interest me. Again, it was only for testing purposes...it is like having a new sports car and you wish to test its abilities...

I think this is healthy, not only for individuals, but for the community.



I wrote I "think" the paper is wrong. And yes, I will try to contact one of the authors of the paper, as I suspect something is amiss here.



I fail to see the difference here. But I will look into it.