15:05:55 <gaba> #startmeeting metrics July 23rd 2020
15:05:55 <MeetBot> Meeting started Thu Jul 23 15:05:55 2020 UTC.  The chair is gaba. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:05:55 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:06:21 <karsten> important dates: onionperf demo next wednesday, current sprint end next thursday.
15:08:05 <gaba> do we have anything for the agenda?
15:08:10 <gaba> mikeperry: i think you have something?
15:08:51 <woswos> I can bring up the captcha monitoring project again since karsten is here this week
15:08:57 <mikeperry> I believe there may have been questions for me before karsten's vacation? like in early july?
15:09:08 <mikeperry> I don't know what those were; sorry for missing that meeting
15:09:21 <gaba> i think it was related to the onionperf instances
15:09:37 * gaba looking
15:09:42 <gaba> woswos: yes please
15:10:00 <moze> This is my first time ever on IRC.But I am the author for today's group paper "Emotional and Practical Considerations towards the Adoption and Abandonment of VPNs as a Privacy-Enhancing Technology".
15:10:05 <karsten> woswos: it's on my list to take a look at that. I'm just not sure when I'll find time to look into non-onionperf things again.
15:10:50 <gaba> moze: welcome to irc and Tor! I'm thinking  you are planning to be in the anti-censorship meeting that starts in 50 min
15:10:53 <gaba> right?
15:12:08 <mikeperry> moze: interesting paper choice!
15:12:29 <mikeperry> are you one of the authors?
15:12:43 <woswos> So, I have been working on a project which aims to measure how often each exit relay gets captcha. I mean, the tor users who are using the specific exit relays faces captcha while browsing websites. I thought we can integrate the data I collect into the metrics website. You can find the data I collected so far here: https://dashboard.captcha.wtf/
15:12:43 <woswos> and API here https://api.captcha.wtf/
15:13:31 <woswos> and I created a ticket to discuss this issue https://gitlab.torproject.org/tpo/metrics/website/-/issues/40002
15:13:53 <karsten> woswos: I haven't looked yet, and I believe it's interesting stuff. I just don't know what to do with it right now or in the next weeks.
15:14:08 <moze> gaba: thank you, yes, I am planning to part of the anti-censorship meeting.
15:14:53 <karsten> woswos: I'm really sorry, but I don't think I can work on that issue in the near future. we have so little capacity in the metrics team right now, it's hilarious.
15:15:01 <woswos> karsten It is not completely ready yet, so I will keep working on it. I wanted to see if metrics people would be interested in such an integration.
15:15:32 <moze> mikeperry: yes, I am the main author.
15:15:34 <woswos> no, I will work on it. I'm not asking you do extra work.
15:16:18 <karsten> okay, sounds good. let's leave the issue open until there's more time, okay?
15:16:36 <gaba> karsten, woswos: it seems to me that an integration would be great!
15:16:40 <karsten> (and maybe ridiculous is a better word than hilarious in this context.)
15:16:41 <woswos> sure
15:16:55 <karsten> gaba: let's keep this in mind when talking about hats.
15:17:01 <gaba> ok
15:17:06 <woswos> great, thanks
15:17:49 <gaba> anything else? do people need help with the stuff on your plate?
15:18:21 <karsten> anything onionperf-related where I can help unblock people?
15:18:50 <gaba> we have a few tickets in the next column https://gitlab.torproject.org/tpo/metrics/onionperf/-/boards
15:20:13 <acute> I had a look at #33399 (in need of review), and would like to test the code changes
15:21:20 <karsten> do you need anything from me for that? I might have local notes that aren't on the issue yet.
15:21:52 <acute> just how to get the version of tor required
15:22:12 <karsten> I'll take a look and will comment on the issue.
15:22:40 <acute> ok, thank you very much
15:22:43 <karsten> sure!
15:22:52 <acute> that's all from me
15:23:09 <karsten> great. thanks for all your work on the various issues!
15:23:42 <karsten> I don't have anything else for today.
15:23:46 <acute> sure! thank you for the prompt feedback
15:24:01 <karsten> :)
15:24:51 <karsten> mikeperry: did you have anything?
15:25:23 <mikeperry> not specifically. do you recall what the questsions you had were, in early july?
15:25:31 <mikeperry> something about the additional onionperf instances, perhaps?
15:26:27 <karsten> I think I might have had questions regarding guards or dropping timeouts, but those issues are a bit on hold right now.
15:26:59 <karsten> I don't recall having questions about additional onionperf instances. did you talk to phw about them?
15:27:53 <mikeperry> oh that's right. I remember now. we were wondering about the monitoring you had for onionperf, and if we can monitor additioonal onionperf instances without necessarily reporting their data to the public website
15:28:58 <karsten> we can run measurements that don't go to the public website, and we can later visualize those measurements using onionperf's built-in visualization code.
15:29:08 <karsten> in fact, that's the plan for most experiments.
15:29:51 <karsten> visulizations would be those long PDFs with dozens of pages. maybe you recall seeing those before.
15:30:20 <karsten> and the data would be a tarball containing raw logs and some analysis files.
15:30:29 <acute> sorry, have to go; bye everyone!
15:30:35 <karsten> bye, acute! o/
15:30:39 <mikeperry> acute: o/
15:30:47 <gaba> o/
15:30:55 <gaba> anything else for today?
15:30:59 <karsten> does that answer the question?
15:31:14 <mikeperry> karsten: ok. will we also have nagios monitoring or similar, to make sure the extra instances are up/reachable/uncensored?
15:31:20 <karsten> ah, yes.
15:31:27 <karsten> as of today (!) we have monit in place.
15:31:36 <mikeperry> oh nice
15:31:53 <karsten> the three instances in nl, us, and hk already monitor themselves and each other.
15:32:09 <karsten> we would simply add those additional instance(s) as remote hosts to their monitoring.
15:33:03 <karsten> most of the credit for having monit running goes to phw and acute, by the way.
15:33:43 <mikeperry> ok great. so phw should be familiar with it, and if it brings any new risks to potentially censored instances
15:34:09 <mikeperry> i think those were my main concerns with the extra instances
15:34:13 <karsten> yes, absolutely, phw knows exactly how monitoring works in this case.
15:34:16 <karsten> cool!
15:34:47 <karsten> okay, anything else for today?
15:35:56 <gaba> cricricri I'm going to end the meeting
15:36:01 <gaba> #endmeeting