15:03:45 <karsten> #startmeeting metrics team
15:03:45 <MeetBot> Meeting started Thu May 16 15:03:45 2019 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:03:45 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:03:54 <karsten> anything else for the agenda?
15:04:14 <irl> not from me
15:04:26 <gaba> not from me
15:04:46 <karsten> great!
15:04:49 <karsten> - bandwidth files update (karsten)
15:05:01 <karsten> so, collector is archiving bandwidth files now!
15:05:13 <irl> woo
15:05:34 <karsten> there's one known bug remaining that affects syncing between collector instances. there's a patch on the ticket.
15:05:46 <karsten> the other remaining part is to do something with tom's archive.
15:05:52 <irl> i'm hoping to deploy that on the test box on monday
15:06:27 <irl> for tom's archive, will the tarballs script do the right thing if we just run them all through collector?
15:06:35 <karsten> not automatically, no.
15:06:59 <karsten> basically, we'd do those tarballs manually.
15:07:08 <karsten> which should be fine.
15:07:11 <irl> we need to add type annotations
15:07:12 <karsten> we're only doing it once.
15:07:29 <karsten> right, so my idea was to make sure the Local mode works with bandwidth files,
15:07:37 <irl> this might actually be a thing to use bushel for, because it has a "collector document" concept which is an encapsulation of a descriptor
15:07:40 <karsten> download them all from tom's server, and make tarballs out of them.
15:07:41 <irl> this would be a good test case
15:08:15 <irl> ah, it's a good test case for local mode too
15:08:44 <karsten> right. how about I try that out, and if it turns out to be more work than I think (for whatever reason) we try bushel?
15:08:48 <irl> and then you're going to make the tarballs by matching with a glob?
15:09:07 <karsten> hm?
15:09:38 <irl> the date is in the filename, so just YYYY-MM-* would work for putting things into monthly tarballs
15:09:45 <irl> with two for loops if you wanted it more automated
15:10:14 <karsten> my current plans for tarballs is to hack create-tarballs.sh a bit.
15:10:23 <karsten> by making tarballs for more than just 2 months.
15:10:43 <karsten> collector would first make sure that files are in the right subdirectory.
15:11:08 <irl> ah, yes, i forgot they are in different directories
15:11:15 <irl> that sounds like an ok plan
15:12:02 <karsten> great!
15:12:15 <karsten> after these two things, we should be done with bandwidth files, right?
15:12:20 <karsten> or did we want to do more?
15:12:38 <irl> i think we would be done with them for now, until some funded work comes along relating to them
15:12:45 <karsten> yes, sounds good.
15:12:55 <karsten> moving on:
15:12:59 <karsten> - OnionPerf update (karsten)
15:13:16 <karsten> I saw ticket updates before/during my previous meeting.
15:13:33 <karsten> I didn't have the chance to look, but this is something I'm planning to do next.
15:13:46 <karsten> would be great to add these graphs soon.
15:14:07 <karsten> question: tor version update?
15:14:19 <karsten> also, available disk space?
15:14:38 <irl> these haven't been looked at yet
15:14:53 <karsten> ok. do you have an idea when they might fit in?
15:15:09 <irl> tuesday?
15:15:20 <karsten> sounds good!
15:15:50 <karsten> that's all on that topic from me for the moment.
15:16:05 <karsten> - Roadmap: how are we doing with the tasks in progress?
15:16:24 <karsten> gaba?
15:16:41 <gaba> hey
15:16:54 <gaba> it seems we are on track on many of the tasks in progress
15:17:13 <karsten> I have a few suggestions for moving cards.
15:17:26 <karsten> #26838 is done, I think.
15:17:28 <gaba> mmm, just a sec that I clicked the wrong thing in storm...
15:17:31 <gaba> yes
15:17:46 <irl> #28322 is in progress
15:18:01 <irl> #26597 is on review
15:18:30 <karsten> right.
15:18:35 <irl> #29366 is done
15:18:44 <gaba> #21378 is the one you were talking about. Is still in progress, right?
15:19:19 <karsten> #21378 is what we talked about, yes.
15:19:26 <karsten> and it's still in progress, but close to completion.
15:19:28 <irl> #23758 should go to the icebox probably?
15:19:33 <irl> whoa no wrong ticket
15:19:42 <irl> #23752 should go to the icebox probably?
15:19:55 <karsten> yes.
15:20:10 <gaba> or backlog? is not coming back any time soon?
15:20:37 <karsten> as time permits.
15:20:40 <gaba> ok
15:21:17 <gaba> the others on in progress?
15:21:24 <karsten> I was wondering if #29653 and #69650 should go into the icebox for now, too.
15:21:30 <karsten> errr
15:21:41 <karsten> #29650
15:22:00 <irl> #29653 should go to icebox, not #29650
15:22:11 <karsten> ok.
15:22:46 <irl> #29624 is the bit that i'm still working on anyway
15:22:52 <karsten> and I was wondering if "Perform one off page load metrics for user perf metrics" could go into backlog and soon into in progress.
15:23:11 <karsten> would be something to discuss at moz all hands.
15:23:25 <karsten> if it happens in the next few weeks.
15:23:30 <gaba> it sounds good
15:24:26 <karsten> what do you think, irl?
15:24:41 <irl> is that something you would work on, or is it for me?
15:25:31 <karsten> well, I could do something there, but it would be good if you reserved at least some time for this.
15:25:40 <irl> my workstation is in a mess at the moment so trying to get a working selenium+tor browser setup might take more time than it's worth
15:26:15 <irl> something broke in an apt upgrade recently and i've not worked out what it was yet
15:26:33 <irl> when would it need to be before?
15:26:56 <gaba> it would be good to have any work done that may help the conversation for the all hands
15:27:05 <karsten> I think starting measurements by end of the month, running them for a week, and then looking at the results would be good.
15:27:42 <karsten> let's move it to backlog and either of us starts working on it?
15:27:56 <irl> right, a week's worth of measurements is not something i could magic up this month
15:28:34 <karsten> okay, happy to grab the card.
15:28:49 <irl> unless, i could run it on aws
15:28:54 <karsten> well, sure.
15:29:05 <karsten> if that works, too, I don't see an issue with that.
15:29:10 <karsten> small machine should be sufficient.
15:29:15 <irl> in which case i should be able to set up something on 28th june
15:29:21 <irl> in which case i should be able to set up something on 28th may
15:29:34 <karsten> or april?
15:29:39 <karsten> may is great!
15:30:02 <irl> i just want to make sure i have a whole clear day, which is why i don't want to say i can do it sooner
15:30:16 <karsten> okay, cool!
15:31:12 <irl> trying out "focussed tuesdays" to follow from "review mondays"
15:31:32 <karsten> heh
15:31:34 <irl> and "meeting thursdays"
15:31:40 <karsten> yeah...
15:31:48 <karsten> okay, so backlog?
15:31:55 <irl> yeah
15:32:12 <gaba> I need to leave in 5 min.
15:32:15 <gaba> sorry
15:32:32 <karsten> these were all my updates for the roadmap.
15:32:40 <gaba> Is all the other things in 'in progress' and 'in review' correct?
15:32:47 <irl> i believe so
15:32:52 <karsten> same here.
15:32:53 <gaba> PRD for metrics portal in review?
15:32:54 <gaba> ok
15:32:59 <irl> still waiting for antonela
15:33:06 <gaba> ok
15:33:21 <irl> not in a huge hurry though
15:33:34 <gaba> one more thing I have is bring the Tor meetup. Does the team needs one space for the morning of one of the days?
15:33:44 <karsten> what do the other teams do?
15:33:58 <karsten> we should probably follow that.
15:34:02 <gaba> still deciding. OONI asked for 1 whole day and it seems that will also be the case for the network team.
15:34:13 <irl> i think we have not run out of roadmap yet
15:34:23 <antonela> irl, what are you waiting from me exactly?
15:34:31 <gaba> https://trac.torproject.org/projects/tor/wiki/org/meetings/2019Stockholm/DailyAgenda
15:34:32 <irl> we could shuffle priorities but we're looking a way down the road maybe it doesn't make sense
15:34:37 <gaba> ohhh, sorry
15:34:38 <karsten> true, but we'll have to make more roadmap for the time until the next meeting.
15:34:38 <irl> antonela: review of the prd for data portal
15:34:46 <gaba> I need to go. I will look at the backlog later. o/
15:34:51 <karsten> o/
15:35:17 <irl> i think we could even have a 1 hour roadmapping session
15:35:22 <irl> if we plan ahead
15:35:23 <karsten> hehe
15:35:45 <irl> in fact, that is what we have on the agenda
15:36:03 <karsten> 1 day sounds like a lot.
15:36:16 <irl> yeah, we could roadmap the next 25 years in that
15:36:22 <antonela> irl: oh, okey
15:36:51 <karsten> my understanding is that if we need less time, that's not an issue.
15:37:01 <karsten> on the other hand we cannot do as much if everyone else is busy with their team.
15:37:19 <irl> i think it may be useful to spy on other roadmapping sessions in the morning
15:37:20 <karsten> maybe we should put this on the agenda for next week?
15:37:33 <irl> one of us can do ooni and the other network
15:37:37 <irl> yeah ok
15:37:44 <karsten> yes, that sounds like a good plan, too!
15:37:53 <karsten> just in case they make plans that involve us.
15:38:12 <irl> exactly
15:38:28 <karsten> cool!
15:38:57 <karsten> I guess gaba will read backlog and would tell us if she needs to know anything else regarding the tor meetup.
15:39:07 <karsten> anything else for today?
15:39:10 <irl> yeah, can follow up with email if needed
15:39:12 <irl> nothing else from me
15:39:14 <karsten> yep.
15:39:39 <karsten> cool! thanks, and bye! o/
15:39:46 <irl> bye!
15:39:52 <karsten> #endmeeting