[tor-project] December 2018 report and January 2019 plans for the metrics team
karsten at torproject.org
Wed Jan 9 16:03:22 UTC 2019
Hello Tor, hello world!
Below you'll find the highlights of Tor metrics team work done in
December 2018 as well as a few expected highlights for the current
month, January 2019.
On behalf of the Tor metrics team,
Published technical report "Towards modernising data collection and
archive for the Tor network"  and implemented a prototype replacement
Published end-of-year campaign blog post "Strength in Numbers: Measuring
Diversity in the Tor Network" .
Updated the "Total consensus weights across bandwidth authorities" graph
to only contain running relays and to also contain consensus numbers [4,
5, 6, 7].
Updated 4 graphs containing OnionPerf data to show all sources
separately rather than an aggregate number [8, 9].
Combined 3 graphs on consumed bandwidth into a single visualization
using stacked area charts [10, 11].
Found a bug  that broke bandwidth history graphs at the end of
November 2018 and that subsequently caused glitches in the user number
graphs, and re-imported missing data from November and December 2018.
Made a couple internal changes to the Tor Metrics website that will make
it a lot easier to maintain in the future, including Java rewrites of
the censorship detector  and parts of the advbwdist module , and
using readr's read_csv() rather than R's save() and load() to speed up
drawing graphs .
Extend OnionPerf to support measuring with a pluggable transport.
Deploy a fourth OnionPerf instance to improve vantage point diversity.
Write a Tech Report that gives an overview of the Tor Metrics codebases.
Use Java 8 date-time functionality  in more Tor Metrics codebases.
Share more code between the modules providing data for Tor Metrics
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 528 bytes
Desc: OpenPGP digital signature
More information about the tor-project