Hi! > I see... > Is the low throughput only caused by the hoster or could it also be the > case that Tor is just not using more of the available bandwidth? > 03/11/11 had an average of 266Mbit/s What CPU is in that (v)Server? > Does a regular download from the server (scp, ftp) confirm the > assumption that the low throughput is caused by the hoster? (excluding > Tor as the problem) Is there a 500M or 1G testfile? > Would it be an option to open the exit policy a bit at that server to > see if that would increase the bandwidth usage? To my knowledge, Tor is not multithreaded, so it's CPU bound to one core. I had it running on my Server with about 270 mbit/s, then one core had 100% util. The exit node was slightly faster (1000 kbyte/s or something) than the blutmagie nodes. The safer solution to utilize more bandwidth is to run 2 or 4 nodes on one (v)server, depending on how much cores you may use (and 2 IPs would be good for 4 nodes, but at the end of the day it doesn't really matter). Best regards! -- morphium - morphium@xxxxxxxxxxxxx - 113332157 http://identi.ca/morphium - http://twitter.com/morphium86