14:30:33 <karsten> #startmeeting metrics team
14:30:33 <MeetBot> Meeting started Thu Nov  1 14:30:33 2018 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:30:33 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:30:43 <karsten> please add any topics that are still missing.
14:31:23 <gaba> https://storm.torproject.org/grain/YYTsbtLcpvGTrBF2Zs4XrJ
14:32:06 <irl> no more topics from me
14:32:32 <karsten> gaba: anything from you?
14:33:30 <gaba> no
14:33:36 <karsten> alright, starting.
14:33:39 <karsten> * Monitoring for operational issues (irl)
14:34:03 <irl> i have been working on writing nagios plugins for monitoring various operational issues. i think we should be able to finish this roadmap item pretty soon.
14:34:12 <karsten> awesome!!
14:34:14 <irl> the first issue addressed already has the new plugin deployed, #28242, so now we know: when an individual onionoo instance is misbehaving and when its varnish is misbehaving. we still test to see that going to onionoo.tpo works, but we only test one of the haproxy frontends when we do this. the architecture for haproxy<->varnish<->onionoo may need some rethinking at some point to actually
14:34:17 <irl> failover nicely so for now i'm considering that check finished.
14:35:10 <karsten> sounds good.
14:35:13 <irl> the next issue is onionperf. i've got a plugin in #28271 that checks to see that the analyze task has been running. we will also check pings, and that the tgen server is running on the internet. i've got some plugins for looking at onion services (that i'm trying to make somewhat reusable) but i've hit a problem:
14:35:15 <irl> https://github.com/robgjansen/onionperf/issues/42
14:35:19 <irl> basically every time you start onionperf you get a new onion address so this makes monitoring pretty difficult. as we're updating the onionperf's tor version etc. anyway, i'm going to see if we can fix this and update at the same time.
14:36:18 <karsten> okay.
14:36:28 <irl> we should have checks in place for collector and for metrics-web too. we want checks that are deeper than just "there is a webserver running" so we should think about what these checks do.
14:36:30 <irl> i wonder if there are some common failure modes we should watch out for.
14:36:51 <karsten> I might already have a list somewhere.
14:36:58 <karsten> I can give that some thoughts next week.
14:37:26 <irl> ok cool. if you can send it to the metrics-team list then i can see it in the archive and give it some thought then
14:37:43 <irl> i will be travelling so won't have email, but can still check the archives on the web
14:37:46 <karsten> do we have a list with current checks?
14:37:58 <irl> the current checks is just the onionoo one
14:38:27 <karsten> I guess I mean that plus the current changes, with considerations what we're monitoring and why.
14:38:34 <karsten> and what we're not monitoring and why.
14:38:50 <irl> ah ok, the two tickets above then have all the details
14:38:54 <karsten> okay.
14:39:48 <irl> that's all i have for this topic
14:40:00 <karsten> cool!
14:40:28 <karsten> * Hackfest in January/February 2019 (karsten)
14:40:37 <karsten> regarding dates,
14:41:20 <karsten> I could arrive on jan 30 in the afternoon and stay until feb 4 in the morning.
14:41:34 <karsten> that's 1/2 wednesday, thursday, friday for the hackfest
14:41:40 <karsten> and saturday and sunday for fosdem.
14:41:57 <karsten> unfortunately, I cannot attend the whole week before fosdem.
14:42:21 <gaba> so it would be only 3 days for metrics. Unless irl and the other person we will have meet before
14:42:40 <irl> i plan to arrive on the 27th
14:42:44 <gaba> ok, let's add that to the pad so jon knows
14:42:45 <irl> and leave on the 4th
14:42:51 <irl> i've already got this on the pad
14:42:56 <karsten> I could stay another day after.
14:43:05 <karsten> but I need to be here on the 29th.
14:43:13 <gaba> ok, np
14:43:15 <irl> i think after fosdem will be not that useful
14:43:20 <karsten> hehe
14:43:25 <irl> i mean, i will be not that useful
14:43:29 <karsten> plausible.
14:43:38 <gaba> is there anybody else you want to invite?
14:43:42 <gaba> that you think should be there
14:43:44 <karsten> our new dev.
14:43:50 <karsten> which we don't have yet.
14:43:51 <gaba> yes
14:43:53 <karsten> who*
14:44:24 <karsten> are there things you could do with network team folks on monday, tuesday, and wednesday morning?
14:44:36 <gaba> yes, they are going to be there the whole week
14:44:58 <gaba> that is other question, interaction with network team
14:45:00 <karsten> right. I wonder if irl could work with them on something cross-team related.
14:45:06 <gaba> they are going to be mostly working on refactoring
14:45:12 <gaba> other than friday that will be roadmap
14:45:48 <gaba> we can have one morning for roadmap retrospective at the metrics team too
14:45:55 <irl> potentially there are controlport things that are useful to metrics that could do with face-to-face interaction with network team
14:46:21 <gaba> teor was suggesting to work on privcount with you both
14:46:34 <karsten> gaba: let's collect possible topics like that one. (the retrospective)
14:46:45 <irl> right ok, i'm happy to look at privcount stuff too
14:46:51 <gaba> https://pad.riseup.net/p/tor-metricsteam-agenda-hackweek-2019.1-keep
14:46:52 <karsten> and maybe the new dev could arrive on the 27th, too.
14:46:53 <gaba> :)
14:47:04 <karsten> :) ok.
14:47:18 <karsten> so, I'll write down 30th to 4th as my dates then.
14:48:22 <karsten> great, that's all from me on that topic.
14:48:28 <irl> i will look at booking travel as soon as i can, to get the best price
14:48:44 <karsten> ah, I'll be driving by car.
14:48:50 <irl> the votes look like we are going to just be in brussels
14:48:52 <karsten> I can take 3 people with me to fosdem.
14:49:05 <karsten> or 2 people and 2 small people.
14:49:27 <karsten> okay.
14:49:34 <gaba> yes
14:49:41 <karsten> even better.
14:49:52 <gaba> we need to look at costs too.
14:50:47 <karsten> regarding the decision whether we're going to brussels or somewhere else?
14:51:27 <karsten> anyway, not something we can discuss today.
14:51:43 <karsten> moving on?
14:51:46 <gaba> yes
14:52:04 <karsten> * Kick off CollecTor Python PoC (karsten)
14:52:11 <karsten> it's a new month!
14:52:25 <karsten> and there's not as much time remaining this year anymore.
14:52:47 <karsten> we need to start working on the collector rewrite soon.
14:53:12 <karsten> I was wondering, given that you're going to be offline next week, irl, whether I can start doing something for the collector deliverable next week.
14:53:26 <irl> yes, i think this will be my big project for once the monitoring is complete
14:53:29 <karsten> I hope to not write as much code for this.
14:53:45 <karsten> but maybe I can prepare this for you to pick up.
14:54:10 <karsten> what's your plan for starting this?
14:54:24 <irl> if you could write a list of requirements for the relaydescs module then that would be a great start
14:54:48 <karsten> I can certainly do that. do you have anything in mind as format?
14:55:08 <karsten> otherwise I'll produce a braindump, but I'm not sure whether that's as useful for you.
14:55:17 <irl> if you can do markdown, eventually this would turn into rST and be documentation
14:55:27 <irl> but markdown is fine for now
14:55:53 <karsten> ok. markdown sounds fine.
14:56:03 <karsten> but what do you expect to read in those requirements?
14:56:32 <karsten> how detailed should they be?
14:56:48 <irl> how often things should be fetched, what documents we should fetch and in what order
14:57:29 <karsten> okay.
14:57:47 <karsten> cool. is on the list.
14:57:50 <irl> thanks (:
14:57:56 <karsten> sure!
14:58:08 <karsten> another thing that came to mind when reading your announcement. ;)
14:58:09 <karsten> * Possible to prioritize #28116 review? (karsten)
14:58:24 <karsten> I was mainly thinking whether I can move that forward while you're away.
14:58:40 <irl> yes, i can make sure this is done before monday
14:59:08 <karsten> okay, that would be really cool. if it doesn't work out, it's okay. if it does, awesome! :)
14:59:31 <karsten> * Monthly report (karsten)
14:59:40 <karsten> I started a draft.
14:59:47 <karsten> let me put it on the pad.
15:00:04 <karsten> there it is.
15:00:46 <karsten> this is less urgent.
15:00:56 <karsten> I just figured that we could get this out this week.
15:01:04 <karsten> but we can as well send it in 1.5 weeks.
15:01:08 <gaba> did you add a link to the draft report somewhere?
15:01:16 <karsten> I pasted it to the agenda pad.
15:01:49 <karsten> do you have it?
15:02:01 <irl> i might be able to look at this on the train on monday if there is not time before then
15:02:16 <karsten> I'd say that's soon enough.
15:02:19 <karsten> later would work, too.
15:02:25 <irl> it is 8 hours of train, so it will probably fit in
15:02:28 <karsten> hah
15:02:40 <gaba> let me refresh it as i dont see it
15:02:52 <karsten> below announcements.
15:03:06 <gaba> ah yes
15:03:37 <karsten> okay, moving on:
15:03:38 <karsten> * Tomorrow's retrospective and roadmap update: open or closed? (karsten)
15:03:57 <karsten> regarding the open/closed part, I think I assumed these were closed.
15:04:22 <gaba> We can have it close. I was not sure if there is somebody else should be there.
15:04:39 <karsten> I can't think of anybody.
15:04:43 <gaba> That is why I was sending both of you a message to see if we send it to the metrics team mailing list or not.
15:04:45 <karsten> I was thinking it's a team-internal thing.
15:04:46 <gaba> ok then
15:04:54 <gaba> In that case we may not have anybody :)
15:04:55 <karsten> can we create a new channel?
15:04:58 <gaba> yes
15:04:59 <karsten> ah, yes.
15:05:05 <karsten> we could do that, too.
15:05:27 <karsten> that is, have the meeting and tell any strangers coming by that it's supposed to be a team-internal meeting.
15:05:53 <karsten> the issue is that we don't have a team-internal mailing list.
15:05:59 <gaba> let's remove the link to the place of the meeting then from the pad
15:06:05 <gaba> and we create one before the meting
15:06:09 <karsten> ok.
15:06:55 <gaba> any question about how we are going to do it?
15:07:08 <karsten> what you sent sounds like a fine plan!
15:07:37 <gaba> ok
15:07:49 <karsten> we're also doing the roadmap update tomorrow, right?
15:08:09 <gaba> yes
15:08:13 <karsten> okay.
15:08:18 <gaba> we can do the roadmap after the retrospective
15:08:28 <karsten> sounds good.
15:08:51 <karsten> that answers my question on this topic. moving on?
15:08:56 <gaba> yes
15:09:05 <karsten> * Timesheet reminder (karsten)
15:09:17 <irl> i have just submitted my timesheets
15:09:21 <karsten> just a reminder that we need october timesheets very soon.
15:09:34 <karsten> including this weeks?
15:09:37 <karsten> week's
15:09:53 <irl> yes
15:09:57 <karsten> okay, perfect.
15:10:02 <karsten> thanks!
15:10:17 <karsten> last topic:
15:10:18 <karsten> * Next meeting time after daylight savings ended (karsten)
15:10:31 <karsten> though, before we talk about that: are we going to have a meeting next week?
15:10:42 <karsten> it sounds like you're away from the internet anyway, irl, right?
15:10:49 <karsten> or just email?
15:11:01 <irl> i would have internet
15:11:05 <irl> but probably can't attend the meeting
15:11:13 <karsten> okay.
15:11:25 <karsten> let's skip next week then.
15:11:28 <irl> i also will be remotely participating in the ietf on bangkok time
15:11:34 <irl> so this will be interesting
15:11:40 <karsten> heh
15:11:53 <gaba> ok
15:11:58 <karsten> for the week after, let's think about shifting time to standard time.
15:12:14 <irl> you mean change the time of the meeting?
15:12:21 <karsten> yes.
15:12:31 <karsten> I have another meeting in 15 minutes.
15:12:45 <karsten> which is going to be moved by 1 hour soon.
15:12:58 <irl> i prefer to not believe in DST
15:13:06 <irl> and not make any changes based on it
15:13:07 <karsten> oh, me, too.
15:13:12 <karsten> I blame the others.
15:13:14 <karsten> ;)
15:13:18 <gaba> :)
15:13:23 <irl> if we make a change then we should make the change that means we don't have to make another
15:13:42 <karsten> then it will collide with the other meeting.
15:13:53 <karsten> next summer.
15:14:01 <irl> next summer europe doesn't have DST
15:14:03 <karsten> still, not having a 1 hour gap would be good.
15:14:08 <karsten> yes, but the states.
15:14:23 <karsten> (maybe I'm confusing things now, this gets far too complicated...)
15:14:50 <karsten> okay, how about this:
15:14:55 <gaba> yes, DST moves all the meetings 1 hour earlier for US :/
15:15:12 <karsten> no meeting next week, meeting at usual time the week after, where we discuss moving or not.
15:15:23 <gaba> it sounds good
15:15:28 <irl> ok\
15:15:52 <karsten> alright.
15:15:57 <karsten> I think we're done.
15:16:00 <karsten> whee.
15:16:19 <irl> (:
15:16:28 <karsten> thanks, irl and gaba! have a good day/evening! bye. :)
15:16:38 <karsten> #endmeeting