|
@@ -1571,10 +1571,11 @@ performance. % Right now the first $500 \times 500\mbox{B}=250\mbox{KB}$
|
|
|
%sendme} acknowledgments arrive.
|
|
%sendme} acknowledgments arrive.
|
|
|
To quantify these effects, we did some informal tests using a network of 4
|
|
To quantify these effects, we did some informal tests using a network of 4
|
|
|
nodes on the same machine (a heavily loaded 1GHz Athlon). We downloaded a 60
|
|
nodes on the same machine (a heavily loaded 1GHz Athlon). We downloaded a 60
|
|
|
-megabyte file from {\tt debian.org} every 10 minutes for 21.3 hours (128 sample
|
|
|
|
|
|
|
+megabyte file from {\tt debian.org} every 30 minutes for 54 hours (108 sample
|
|
|
points). It arrived in about 300 seconds on average, compared to 210s for a
|
|
points). It arrived in about 300 seconds on average, compared to 210s for a
|
|
|
direct download. We ran a similar test on the production Tor network,
|
|
direct download. We ran a similar test on the production Tor network,
|
|
|
-fetching the front page of {\tt cnn.com} (55 kilobytes): while a direct
|
|
|
|
|
|
|
+fetching the front page of {\tt cnn.com} (55 kilobytes) every 10 minutes for
|
|
|
|
|
+21.3 hours (128 sample points): while a direct
|
|
|
download consistently took about 0.3s, the performance through Tor was highly
|
|
download consistently took about 0.3s, the performance through Tor was highly
|
|
|
variable. Some downloads were as fast as 0.3s, with a median at 2.6s, and
|
|
variable. Some downloads were as fast as 0.3s, with a median at 2.6s, and
|
|
|
90\% finishing within 6.0s. It seems that as the network expands, the chance
|
|
90\% finishing within 6.0s. It seems that as the network expands, the chance
|
Nick Mathewson