14:59:35 <karsten> #startmeeting metrics team meeting
14:59:35 <MeetBot> Meeting started Thu Jul  2 14:59:35 2020 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:59:35 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:59:38 <karsten> https://pad.riseup.net/p/tor-metricsteam-2020.1-keep <- agenda pad
14:59:42 <dennis_jackson> o/
14:59:43 <acute> hi!
14:59:48 <phw> o/
14:59:51 <karsten> hi dennis_jackson and acute! and phw!
15:00:19 <karsten> today's roadmap time!
15:00:35 <karsten> I added a few more topics to the agenda. please add more things you want to talk about today.
15:01:58 <karsten> gaba: you around?
15:02:32 <gaba> i am
15:02:37 <gaba> half here but I am :)
15:02:38 <karsten> great! hi!
15:02:41 <gaba> hi !
15:03:00 <karsten> we can start with the other topics if you want.
15:03:06 <gaba> thanks
15:03:07 <gaba> yes
15:03:10 <karsten> okay!
15:04:00 <karsten> alright, let's start.
15:04:04 <karsten> Shutting down op-??2 instances
15:04:14 <karsten> any reason to still keep them running?
15:04:23 <karsten> we have op-??3 instances running for a few weeks now.
15:04:37 <karsten> https://metrics.torproject.org/torperf.html
15:04:54 <karsten> those op-??2 instances were the ones with lots of 50kib and 1mib requests.
15:05:16 <dennis_jackson> None from my end, the overlap looks very reassuring
15:05:24 <karsten> indeed.
15:05:30 <acute> yes
15:05:58 <karsten> okay, great. if I don't hear anything by the end of this meeting, I'll pull the plug.
15:06:23 <karsten> next topic:
15:06:24 <karsten> Merging develop into master and tagging next version
15:06:36 <karsten> so, today we finished a 4 week roadmap.
15:06:47 <karsten> or, are going to finish that at the end of the day.
15:06:58 <karsten> is now a good time to merge the develop branch into master and tag a version?
15:07:08 <gaba> It seems good to me
15:07:49 <acute> I think so too
15:07:52 <phw> seems like a good time to me
15:08:14 <karsten> okay, cool!
15:08:20 <karsten> this model is still new for me.
15:08:33 <karsten> will do that at the end of (my) today.
15:08:55 <karsten> I added another topic:
15:08:55 <karsten> Merge requests
15:09:15 <karsten> I tried to create a merge request but failed.
15:09:22 <karsten> anyone being able to give me a hand after the meeting?
15:09:33 <gaba> karsten: was there an error?
15:09:35 <phw> karsten: did you do it from a repo that you forked?
15:10:03 <phw> (i didn't work for me when i "manually" forked a repo. i had to actually click the fork button)
15:10:08 <karsten> phw: I think I failed to upload my branch to the repo I forked.
15:10:31 <karsten> https://gitlab.torproject.org/karsten/onionperf
15:10:35 <karsten> this is my repo.
15:10:51 <karsten> but I failed to push to it or even fetch from it.
15:11:00 <karsten> something related to permissions, keys, you know, crypto.
15:11:31 <karsten> do you have a minute after the meeting to look into this together with me?
15:11:52 <acute> did you create an access token?
15:12:20 <karsten> no. I did create an ssh key though and put it into gitlab.
15:13:26 <karsten> do I still have to create an access token?
15:14:07 <gaba> I do not think you have to create an access token for forking
15:14:17 <gaba> or doing a MR
15:14:51 <phw> i wonder if the project settings have some overly restrictive options set.
15:15:06 <acute> ok just tested this with my key also, it seems to work
15:15:29 <acute> the only difference was that I cloned the repo with ssh, not with https
15:15:53 <karsten> I think I did that, too.
15:16:06 <karsten> % git remote show karsten
15:16:07 <karsten> git@gitlab.torproject.org: Permission denied (publickey).
15:16:19 <karsten> not sure how I tell it which ssh key to use.
15:16:26 <karsten> I though it'll figure that out by itself.
15:16:44 <karsten> but, I don't want to use too much of this meeting time on this topic.
15:16:55 <acute> does your key appear when in the output of 'ssh-add -L'?
15:17:26 <karsten> no!
15:17:30 <acute> aha!
15:18:30 <karsten> better!
15:18:35 <karsten> thanks so much. :)
15:18:58 <acute> cool!
15:19:07 <karsten> okay, I'll give merge requests another try next time.
15:19:17 <karsten> Roadmap for July
15:19:20 <karsten> finally!
15:19:22 <gaba> yes
15:19:38 <karsten> do you want to go through this?
15:19:39 <gaba> mostly people need to check that they are ok with what is in the next and doing columns
15:19:50 <gaba> with the stuff you all have on your plate now and the capacity you have for next month
15:20:48 * phw is ok with "doing"
15:21:09 <gaba> there are a few issues in Next that I would like to bring. The #28271 is one that maybe phw you can take for next month?
15:22:04 <karsten> regarding that ticket: it doesn't have to be nagios.
15:22:16 <karsten> we just need something to monitor our op-??? instances.
15:22:25 <phw> i was wondering about that, karsten
15:22:40 <phw> monitor that the onionperf process is still running?
15:22:42 <karsten> right now I have a local script on my laptop that fetches the latest log files, greps them for the last heartbeat, and displays those to me.
15:22:55 <karsten> some basic checks.
15:23:17 <karsten> still running and doing measurements.
15:23:40 <karsten> (not sure if there are zombie states where the process keeps running but isn't doing anything useful. we'd like to distinguish those.)
15:23:54 <phw> we have been using the monit tool to monitor anti-censorship infrastructure and i've become quite comfortable with it. i can summarise on the ticket how it could be helpful here
15:24:18 <karsten> are you running (hosting) that?
15:24:22 <phw> and then we can decide how we would like to move forward
15:24:23 <karsten> or where is that running?
15:24:54 <phw> karsten: it's running on my (personal) vps, for remote monitoring. we could also set one up on the monitored machines, and configure the instances to email us if an onionperf breaks
15:24:58 <karsten> or is it something that's going to run on one of the numerous solaris hosts we're going to have soon?
15:25:01 <karsten> ;)
15:25:32 <acute> karsten: the tgen 1.0.0 analysis output can also include the tgen hearbeats, not sure if we can use that for something
15:25:53 <gaba> acute: how are you doing with the stuff in your plate?
15:26:07 <karsten> acute: but those would come with a 24 hour delay, right?
15:26:23 <acute> karsten: yes
15:26:28 <karsten> but okay, not going too deep. phw: having your thoughts on that ticket would be cool!
15:26:32 <karsten> err, issue, not ticket.
15:26:48 <acute> gaba: not expecting 33974 to take much longer :)
15:26:49 <karsten> acute: ideally, we'd learn sooner.
15:26:50 <phw> aye aye, i will comment
15:26:54 <karsten> thanks!
15:27:12 <gaba> great!
15:27:16 <karsten> 33974 is quite a beast.
15:27:40 <gaba> acute: if you have bandwidth next month to any other TGen ticket then go ahead :)
15:27:41 <karsten> do you need any help with that, in particular the visualization update part?
15:28:13 <acute> not at the moment, thanks for offering!
15:28:29 <acute> if would be good to have another pair of eyes on the analysis output however
15:28:55 <karsten> when you have something, let me know.
15:29:05 <acute> will update the ticket!
15:29:09 <karsten> cool!
15:29:27 <karsten> acute: regarding #30362, do you have any idea what remains to be done after #33974 is in?
15:31:05 <karsten> no need to answer this now. but can we add a note to #33974 to update #30362 as last step?
15:31:17 <acute> so, the new tgen parser calculates the start times of transfers in a similar way to the old one
15:31:23 <karsten> ah, okay.
15:31:26 <karsten> so the issue remains.
15:31:37 <karsten> except that we'd want to fix it upstream?
15:31:38 <gaba> karsten: is your plan to take the s55 tickets in the next couple of weeks, right?
15:31:47 <karsten> gaba: yes!
15:31:48 <acute> yes, I think so
15:32:06 <karsten> I'll quickly add that to the ticket.
15:32:09 <karsten> issue*
15:33:35 <karsten> done.
15:33:56 <acute> thanks!
15:34:10 <karsten> gaba: so, yes, I'll be distracted a bit by s55 and a couple other smaller things.
15:34:18 <gaba> ok
15:34:20 <gaba> sounds good
15:34:27 <karsten> gaba: I still plan to work on onionperf tickets.
15:34:48 <karsten> looking at the board, did we say this is for 4 weeks or 8?
15:35:05 <karsten> (gaba and I briefly talked about this on monday already.)
15:35:13 <gaba> 4 weeks
15:35:26 <karsten> what remains for the 4 weeks after?
15:35:48 <gaba> There is still stuff in the backlog
15:36:07 <gaba> Altough we said is not required for this project.
15:36:12 <karsten> yeah...
15:36:34 <karsten> hmm.
15:37:07 <karsten> I mean, I guess we can try how far we get.
15:37:19 <karsten> the issues in Next, Doing, and Needs Review all look important to me.
15:37:25 <gaba> yes
15:37:29 <gaba> let's focus on those
15:37:35 <gaba> and we see how we are doing at the beginning of August
15:37:59 <karsten> are we missing anything, acute, phw, dennis_jackson, jnewsome?
15:38:17 <karsten> mikeperry, if you're around?
15:38:38 <phw> i'm good
15:38:52 <acute> this looks good
15:39:15 <karsten> okay, great!
15:39:29 <dennis_jackson> :+1:
15:39:48 <karsten> acute: having yet another comment from you on #40001 would be neat. then I could include the updated README.md in master later today.
15:40:02 <karsten> thanks for all the reviews, by the way! much appreciated. :)
15:40:22 <acute> excellent! will update just after the meeting
15:40:28 <karsten> perfect!
15:40:50 <karsten> anything else on the roadmap topic?
15:41:00 <gaba> not from me
15:41:04 <jnewsome> karsten: sorry, where are we looking? I'm kind of just keeping one eye on this
15:41:15 <karsten> jnewsome: https://gitlab.torproject.org/tpo/metrics/onionperf/-/boards
15:41:43 <karsten> jnewsome: anything missing from Next, Doing, Needs Review that we should be doing in the next 4 (or 8) weeks?
15:42:45 <karsten> speaking of, if you'd like to look at the TGen 1.0.0 stuff that is pretty far by now, feel free to do that. more eyes == better
15:43:32 <jnewsome> thanks; ok, i'll try to take a look
15:43:37 <karsten> cool!
15:43:43 <karsten> anything else for today?
15:44:05 <gaba> i'm fine
15:44:11 <dennis_jackson> all good here
15:44:13 <acute> same
15:44:48 <karsten> great! alright, let's end the meeting and get back to work. thanks, everyone! bye. o/
15:44:55 <jnewsome> thanks! o/
15:44:56 <gaba> o/
15:45:00 <acute> bye!
15:45:05 <karsten> #endmeeting