18:59:36 #startmeeting network health 18:59:36 Meeting started Mon Feb 17 18:59:36 2020 UTC. The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:59:36 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:59:45 alright everyone 18:59:55 hi 19:00:03 let's do some network health meeting this 19:00:05 week 19:00:08 dennis_jackson: o/ 19:00:12 hello 19:00:19 our meeting pad is at: https://pad.riseup.net/p/tor-networkhealth-2020.1-keep 19:00:35 if you have things to add to the status section, now would be a good time :) 19:00:49 the same goes for the discussion section 19:01:03 o/ 19:01:41 * ahf is here with half an eye but is cooking too 19:02:47 okay, let's get started 19:03:04 one sec 19:03:10 k 19:04:28 done 19:04:46 ahf: i saw your churn rate experiments, good stuff 19:04:51 keep them going :) 19:04:57 okay 19:05:17 there is a next step in it where i hope to visualize things a bit better. will report back when i have something 19:05:20 wont be this week though :-/ 19:05:38 so, just as a reminder we have a bunch of things already tagged for our roadmap in the coming weeks 19:05:58 if you look for the keyword `network-health-roadmap-2020Q1` it is 19:06:18 it's the most reliable way to find out about tickets on the radar right now 19:06:33 if you tag them with 'network-health-roadmap-2020Q1' then they will up in the network health team wiki page 19:06:33 as we don't have a dedicated network-health component (yet) 19:06:36 https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkHealthTeam 19:06:46 yep 19:06:56 thanks for that, gaba 19:08:04 dennis_jackson: i saw your work on #33076, very nice 19:08:17 I think this component is more metrics team than community team - https://trac.torproject.org/projects/tor/ticket/33175 19:08:29 Thanks Geko! 19:08:41 so, if you want to think about potentially useful things for a network health perspective (aka #33178) while doing that work 19:08:47 that would be much appreciated 19:09:13 i plan to start in a bit pondering that part, too, but more folks thinking about that can't hurt :) 19:09:31 ggus: yeah 19:09:33 :) indeed. If I come up with something cohesive, I'll be sure to share it 19:09:38 thx 19:11:00 ggus: i changed it, thanks 19:11:45 there is nothing marked bold at the status updated for discussion 19:11:55 so, let's move on to discussion 19:11:58 ggus: you are up 19:12:18 ok 19:12:38 some weeks ago we discussed in #tor-project the next steps for strengthen relationships with non-profit orgs that run relays 19:12:51 i would like to discuss here too 19:13:07 i was hoping that arma2 would be available for this discussion 19:13:16 but it's holiday in us 19:13:56 well, we could start discussing it here and now 19:14:11 or we could postpone it to next week if you thinkg it's move useful that way 19:14:17 *think 19:14:28 maybe both could be helpful :) 19:14:44 in that we can think over the week about some ideas and finish discussion next week 19:14:46 let's start today so, next week i will be offline on monday 19:14:49 but, not sure 19:14:56 aha, okay 19:15:05 let me find a ticket 19:16:05 https://dip.torproject.org/torproject/web/community/issues/63 19:16:28 6 month ago i started to do this table based on torservers.net 19:17:54 there is a discussion now in tor-relays about one single person running 21% of the relays. Are they one of those orgs in your table? 19:17:57 https://lists.torproject.org/pipermail/tor-relays/2020-February/018154.html 19:18:23 gaba: nope, niffty wasn't a torserver partner 19:19:13 *21% of exit node capacity and not 21% of the relays. 19:19:22 21% of the relays is more... dramatic 19:20:04 oh right 19:20:05 yes 19:21:14 ggus: what's the goal with this list compared to "a list of relays operators we know/are in contact with"? 19:21:48 that is what's the goal with the focus on asscociations? 19:21:49 GeKo: this list is going to be in community.tpo with a link to donation 19:21:56 *associations 19:22:32 okay, so the idea would be to have a means to strenghten those orgs 19:22:35 GeKo: where i can find network health roadmap? 19:22:53 via donations coming from the community portal 19:23:02 also that 19:24:11 ggus: https://pad.riseup.net/p/network-health-team-2020Q1-roadmap-temporal 19:24:16 probably 19:24:31 not sure if we have a better write-up besides that one and the tickets 19:24:34 DEPRECATED PAD - GABA COPIED CONTENT TO FILE IN NEXTCLOUD - ASK HER IF YOU NEED TO CHANGE THIS PAD 19:24:49 gaba: where i can find the roadmap? 19:25:08 ok, i found what i needed 19:25:13 ok 19:25:20 i linked the spreadsheet there i think 19:25:26 but everything is in tickets in trac 19:25:35 the spreadsheet was for me to organize myself 19:25:47 task #1 would be to map all these orgs in community portal (we have a ticket open for this); task #2 would be to draft a diagnose form to learn how they are doing, what kind of help they need 19:25:50 ahh, the content of the pad was copied to nextcloud only to archive it 19:27:15 ggus: is that still the plan you outlined there? 19:27:31 that is your #1, #2, and #3? 19:27:43 yes 19:28:44 okay, sounds good to me 19:29:05 how do you envision this is going to happen? 19:29:18 like who is doing what/or should do what? 19:29:39 i can draft the form, people can review it, and i can send the emails to the orgs 19:30:53 if it is appropriate, it might be interesting to gather some background on the relays themselves. I do not know if it would have any security/privacy ramifications though 19:31:14 E.g. "My relay sees less traffic than I think it should?" 19:31:23 (True/False or whatever) 19:31:31 It would be great to get some orgs in the south to run relays 19:31:33 "My relay appears to be CPU bound?" 19:31:48 dennis_jackson: good idea 19:33:08 ggus: sounds like a good plan 19:33:43 depending on how much effort we want to do during the first contact we could add questions like the ones dennis_jackson raised 19:33:55 or we could do those on a follow-up if we wanted 19:34:06 Yeah, maybe it makes sense to build the relationship first 19:34:13 yep 19:34:28 and let the folks speak first and pick things up from there 19:34:53 but it could be good to do some research in parallel 19:35:08 as it could help showing that we really care here 19:35:11 (which we do) 19:35:52 ggus: what do you think? does that work for you? 19:36:25 GeKo: yes, i can start an initial convo and then we submit the questions. 19:36:51 okay. what timeframe to you have in mind for each of those steps? 19:38:53 I'd like to have a form done in march, and the results compiled late in April, to discuss this in IFF Tor Village 19:40:18 sounds good 19:40:57 if you have specific network health things i/we should do that come to your mind while working on this, just ping me 19:41:42 okay, anything else for discussion today? 19:41:58 requests/concerns etc.? 19:42:08 GeKo: i'll need help on the technical part of this. like do we want to know how these orgs manage their relays (like with Ansible)? or what resolvers they use? 19:43:15 yeah, i'll think about those (and i think we'd like to know about those two things you mentioned at least) 19:44:42 ok :) 19:44:52 okay, i think we are done here for today 19:45:02 thanks everyone and have a nice week! 19:45:07 #endmeeting