[tor-dev] Network Health team and analysis work

[Georg Koppen]

Hello everyone!

The Network Health team is best known for its work in the bad-relays 
area and being concerned with providing metrics + keeping an eye on the 
health of the Tor network. While that involves doing analyses to answer 
our own questions it was not clear so far what we should do with 
questions about the network or available data which got raised by other 
teams (i.e. you). Should we do those analyses as well or should we just 
provide the necessary data and each team then tries to answer their 
questions by themselves? This got raised a bunch of times in the 
past[1][2] and I am happy to announce that the Network Health team is 
stepping up and helping you with your analyses, if desired.

The idea would be that we'd do the analysis work in those cases, alone 
or together, and accumulate/improve a set of scripts and tools over time 
at a single place, so that both we and other interested parties can do 
those analyses faster and better in the long(er) run. The tools should 
be open so that other folks can help improving and shaping them 
according to their needs.

We have a dedicated analysis project[3] where we track all analysis 
related issues and which contains tools and documentation about how to 
use them for our work. If there is any analysis work you would like us 
to get done, please file a ticket in that project. If you did a similar 
analysis in the past by yourself we'd love to hear about it as it might 
speed up our work and we could build documentation out of it and add 
available code to the analysis project. Additionally, please assume we 
are no experts in your area of work, but don't hesitate to file tickets 
because of that. We'll ask if things are unclear. :)

To get your analysis questions answered in a timely manner it would 
greatly help if they were attached to some sponsored work as that makes 
it usually easier to justify picking them up among all the other things 
on our plate. Additionally, emergencies (like new blocking events in 
country X) should be easily justifiable as well, in particular as we 
have a "surprise 'anomaly analysis' on the network as needed"-item on 
our roadmap throughout the year. I plan to make sure that we have 
sufficient spare cycles available for that work should it be needed. If 
your desired analysis does not fall into either of those categories 
don't despair, we'll try to find a way to get it done nonetheless.

Thanks,
Georg

P.S.: If you are not part of any of our teams but want to do an analysis 
by yourself or just note an analysis idea somewhere, feel free to start 
by filing a ticket in the analysis project as well. Others might just 
jump on it and get it done because they might be curious about the 
answers themselves. :)

[1] 
https://gitlab.torproject.org/tpo/network-health/metrics/relay-search/-/issues/40019
[2] 
https://gitlab.torproject.org/tpo/network-health/team/-/issues/250#note_2863179
[3] https://gitlab.torproject.org/tpo/network-health/analysis
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20230706/ee6d3533/attachment.sig>