- tor-relays - lists.torproject.org

How to fully utilize resources for relay?
by Steve Snyder 05 Aug '11

05 Aug '11

I have a machine that is dedicated to being a Tor exit node. How can I maximize the performance (high throughput / low latency) of Tor traffic? I have plenty of under-utilized CPU and RAM resources on this system. According to TorStatus, my bandwidth is rarely maxed-out. This is on a Linux system, currently running Tor v0.2.2.30-rc. So... what resources can I spend to make this a better Tor node? Thanks.

3 3

Number of Connections Open
by Tim Sally 02 Aug '11

02 Aug '11

Hi! I'm working on a model to estimate the number of connections an OR has open as a function of the OR's bandwidth. I could really use some help confirming that my model is reasonable. Would any relay operators be willing to share the results of some variant of netstat | wc -l and the bandwidth of the relay? Thanks, Tim

8 7

Flags
by vinc3nt 01 Aug '11

01 Aug '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 hello, I have set up a Tor relay. Where can I find online documentation that explains the meaning of all flags set: Authority Bad Directory Bad Exit Exit Fast Guard Hibernating Named Stable Running Valid V2Dir Some are intuitive, but others do not understand the meaning tnx - -- vinc3nt +--------------------------------------------------------------------+ |GPG Id key: 4096R/1E2A7F06 2010-11-23 http://pgp.mit.edu | |Key Fingerprint: 1355 B43E 99A9 37EF 5F27 A399 B71C 8AEB 1E2A 7F06 | |() ascii ribbon campaign - against html e-mail | |/\ www.asciiribbon.org - against proprietary attachments | +--------------------------------------------------------------------+ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) iQIcBAEBAgAGBQJONoYsAAoJELcciuseKn8GEX4P/jmXuLHL7nKyDA7vOegKMeii HgesWX2cUbxpVkOzbVZ/ubtE2RMFmVfikS4ChbcIFzAECOKRctY1lDy45Sihwtit jNysYwm9Fh/3uBlpRPMdMhHA8g7rr6Xh5lDvkSTdeyJVAx0xnOLNEhT3Oi2aYsZF a2vgcwTLdsvgDRxMynDhDKwxKIfyPZk443gEYMaT3kxVlxr18zD8XDsFxbVOl9LO HhCZ3USJYHW/rWf3v1+HsIx28CHl0RaxPkwtMzX9oXvN5mLYyYy1O0w2wV7mDnp3 59BOAhxbe3oY6F4fmBrw3FyPfgUDPfD0DaY6UZxnokXeRG743XK8AkfXTkiP0g4B /MxBqWypqcNGUvlToii7GBW8RZt+wBukeS50Qok8+RLIIB7SZSPVx90GnZxCl2q5 tnrlqM0dNKi0/jXUHpcmuxsYVLGQ+k+8Ky2E106pagYiayYeajbhuXjPOmZIqYMu dXTqotR1p4JbJFnrmyFRomgDetKUBaZRk6Kjnln8grQyMglPoKFb/1EvtgxrKMOf hysEakFhIAV8yeCYiQAPm0vfSudgpz6i6reStqTMVrjQqm/HObedSTvN+h/Oy2QX MR1r8TTOSdRZgglxy59g/H4EruQxSET5P/RdnCBcRbLp6xDymPOilpSl+bhBS8KO zOt0C+asU/krtS/XDVzt =KUbN -----END PGP SIGNATURE-----

2 2

stable flag
by Softail 25 Jul '11

25 Jul '11

My relay seems to gain and lose the stable flag every day for the last several days. It's been up more than a week. Is this normal? I don't recall seeing this on a previous hosting provider and I suspect there is something going on with the network. I'm thinking of setting the log level to info or debug to try to figure out what is going on. Is that a good idea? Is it likely to be informative? Thanks -- A man in chains knows he should have acted sooner... Julian Assange

3 4

possible bad exit or bad circuit
by cmeclax-sazri 25 Jul '11

25 Jul '11

jceaovh Pandora18 tornodeca I was browsing a website that I hang out on all the time, and it came up with a Privoxy forwarding failure error. I tried another website and got the same error. I restarted Privoxy; didn't fix it. I brought up Vidalia and looked at the circuits. The one listed above was first; when I attempted to view the website, it added a line with the name of the website, labeled Closed, which quickly disappeared. I closed the circuit and refreshed the website and it came up.

2 5

Strange warnings
by Klaus Layer 22 Jul '11

22 Jul '11

Hi, My relays suddenly show the following warnings: "Tried to establish rendezvous on non-OR or non-edge circuit." Is this something I should worry about? I am wondering why several of my relays shows that at nearly the same time. Thanks, Klaus

2 1

Odd activity on my rely
by George Gemelos 19 Jul '11

19 Jul '11

I'm running an exit node, yreka and have noticed as single IP address that has been connected for at least a day, probably longer, which is pushing anywhere from 3Mb to 9Mb but receives very little traffic back from my node, sub 500Kb. My traffic to this machine is going out of 443 (my ORPort) and being sent to port 50090 on the other machine. I only allow exit traffic to ports 80 and 443. So I do not think this is exit traffic. I have also checked on blutmagie and could not find the IP as a relay which would suggest that this is a TOR user and I am serving as an entry node. I have never seen such speeds for a single user. The one sided nature of the traffic and its size have my interests peaked. I just wanted to verify that this is not unusual and something that I should investigating. I did not include the IP address of the machine because I was not sure it would be in the spirit of TOR to post such information. Any feedback would be greatly appreciated.

4 8

Seeking headless TAILS (or similar)
by Steve Snyder 19 Jul '11

19 Jul '11

I'm looking for a Tor installation that boots off of read-only media and acts as a relay, sort of a TAILS but without the user-space applications. Does such a thing exist? Thanks.

2 2

control_auth_cookie problem
by mick 18 Jul '11

18 Jul '11

Hi Guys I use munin to monitor a bunch of stuff (including tor) on my servers. The long run munin stats are quite useful. I checked my stats today and noticed that there were no entries for tor on one machine going back a couple of weeks. Now I know I rebooted the machine in question about 12 days ago and I /think/ I upgraded tor at that time. Certainly the restart will have caused a regeneration of the cookie file anyway. Checking my logs, I see the following from munin: "Unable to connect to Tor ControlPort (515 Authentication failed: Wrong length on authentication cookie" and similarly in the tor log: " Got authentication cookie with wrong length (16)" Now I can see that the cookie file is 32 characters long, so I'm guessing that the file contains a control character that is screwing up the munin php plugin line "$cookie = file_get_contents($cookiepath);" Anyone else seen this before? Any ideas for fixing it? - short of restarting tor again (I've done that and I'm reluctant to keep on stopping and starting what should be a stable server), Thanks Mick --------------------------------------------------------------------- The text file for RFC 854 contains exactly 854 lines. Do you think there is any cosmic significance in this? Douglas E Comer - Internetworking with TCP/IP Volume 1 http://www.ietf.org/rfc/rfc854.txt ---------------------------------------------------------------------

2 1

DNAT question
by Softail 18 Jul '11

18 Jul '11

I'm trying to switch my ports from 9001/9030 to 443/80. The tor configuration seems straight forward. I tried /sbin/iptables --append INPUT --protocol tcp --match state --state NEW --destination a.b.c.d --dport 9001 --jump ACCEPT /sbin/iptables --append INPUT --protocol tcp --match state --state NEW --destination a.b.c.d --dport 9030 --jump ACCEPT /sbin/iptables --table nat --append PREROUTING --protocol tcp --source-port 443 --destination a.b.c.d --jump DNAT --to-destination :9001 /sbin/iptables --table nat --append PREROUTING --protocol tcp --source-port 80 --destination a.b.c.d --jump DNAT --to-destination :9030 but that doesn't seem to work. The OR and Dir ports are not reachable from the outside. I assumed that PREROUTING happened before INPUT but not really an expert on this. The firewall blocks everything else inbound to that address but the two ports I opened. Do I need to open 443/80 on the INPUT chain as well and if so do I also need to keep 9001/9030 open also? CentOS 5.6 if that matters. Thanks -- A man in chains knows he should have acted sooner... Julian Assange

2 4