[metrics-team] Measuring bwscanner changes

Philipp Winter phw at nymity.ch
Wed Jun 7 15:09:56 UTC 2017


On Tue, Jun 06, 2017 at 07:02:18PM -0400, micah anderson wrote:
> The theory is that Bandwidth Authority location matters because Tor
> sends more Guard and Middle bandwidth to relays close to the bandwidth
> scanner, and more Exit bandwidth to relays close to the bandwidth
> server.

Doesn't the scanner's location affect latency much more than throughput?
That is not to say that the effect on throughput is negligible -- longer
paths are more likely to be congested -- but I wonder what the exact
numbers are.

We could use the RIPE Atlas network to learn more about latency
variations (using pings and traceroutes) but there's no easy way to use
Atlas for throughput measurements.


More information about the metrics-team mailing list