[13:10:26] <grrrit-wm>	 (PS1) Erik Zachte: protect against dounble run with flock [analytics/wikistats] - https://gerrit.wikimedia.org/r/126240
[13:10:41] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] protect against dounble run with flock [analytics/wikistats] - https://gerrit.wikimedia.org/r/126240 (owner: Erik Zachte)
[13:16:19] <grrrit-wm>	 (PS1) Erik Zachte: 'incl bots' -> 'excl bots' [analytics/wikistats] - https://gerrit.wikimedia.org/r/126242
[13:16:24] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] 'incl bots' -> 'excl bots' [analytics/wikistats] - https://gerrit.wikimedia.org/r/126242 (owner: Erik Zachte)
[14:59:11] <grrrit-wm>	 (PS7) Milimetric: Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[14:59:20] <grrrit-wm>	 (CR) Milimetric: Fix parse_username to handle Unicode (5 comments) [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[14:59:30] <grrrit-wm>	 (PS8) Milimetric: Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:08:36] <grrrit-wm>	 (PS1) Erik Zachte: massive cleanup of comscore files (standard naming now) [analytics/wikistats] - https://gerrit.wikimedia.org/r/126262
[15:08:43] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] massive cleanup of comscore files (standard naming now) [analytics/wikistats] - https://gerrit.wikimedia.org/r/126262 (owner: Erik Zachte)
[15:08:56] <grrrit-wm>	 (CR) Csalvia: [V: 2] Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:10:36] <grrrit-wm>	 (CR) Nuria: [C: 1 V: 2] Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:11:03] <grrrit-wm>	 (CR) QChris: "Jenkins is complaining, as the test suite was been moved away." [analytics/wikistats] - https://gerrit.wikimedia.org/r/126262 (owner: Erik Zachte)
[15:14:50] <grrrit-wm>	 (CR) QChris: "recheck" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126262 (owner: Erik Zachte)
[15:15:10] <grrrit-wm>	 (PS1) Erik Zachte: use new comScore csv file names [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264
[15:15:15] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] use new comScore csv file names [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[15:16:06] <grrrit-wm>	 (PS1) Erik Zachte: handle missing values slightly different [analytics/wikistats] - https://gerrit.wikimedia.org/r/126265
[15:16:11] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] handle missing values slightly different [analytics/wikistats] - https://gerrit.wikimedia.org/r/126265 (owner: Erik Zachte)
[15:25:08] <grrrit-wm>	 (CR) QChris: "recheck" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[15:25:10] <grrrit-wm>	 (CR) QChris: "recheck" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126265 (owner: Erik Zachte)
[15:28:06] <grrrit-wm>	 (CR) QChris: "Jenkins does not seem to pick up the "recheck"." [analytics/wikistats] - https://gerrit.wikimedia.org/r/126262 (owner: Erik Zachte)
[15:29:21] <grrrit-wm>	 (CR) QChris: "Jenkins does not seem to pick up the "recheck". But when I tell Jenkins to rebuild directly, it goes green:" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[15:30:22] <grrrit-wm>	 (CR) QChris: "Jenkins does not seem to pick up the "recheck". But when I" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126265 (owner: Erik Zachte)
[15:32:14] <qchris>	 hashar: Can I get Jenkins to revote on a change (without uploading a new patch set) if it does not pick up "recheck"?
[15:32:28] <qchris>	 https://gerrit.wikimedia.org/r/#/c/126264/1
[15:34:08] <grrrit-wm>	 (PS9) Milimetric: Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:34:39] <grrrit-wm>	 (CR) Milimetric: Fix parse_username to handle Unicode (1 comment) [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:34:46] <grrrit-wm>	 (CR) Csalvia: [C: 2 V: 2] Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:35:11] <grrrit-wm>	 (PS1) Erik Zachte: dataset2 -> dataset1001 + remove test code [analytics/wikistats] - https://gerrit.wikimedia.org/r/126371
[15:35:43] <grrrit-wm>	 (CR) Nuria: [C: 2] Fix parse_username to handle Unicode [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/125752 (owner: Csalvia)
[15:38:45] <YuviPanda>	 milimetric: +1 for postgres in eventlogging
[15:38:55] <YuviPanda>	 milimetric: but cross dbengine joins :(
[15:38:56] <milimetric>	 seriously YuviPanda :)
[15:39:08] <milimetric>	 yeah, no, we need a postgresql cluster
[15:39:12] <milimetric>	 that takes in everything
[15:39:22] <milimetric>	 Aaron S. and I were talking about this last time I was in SF
[15:39:29] <milimetric>	 there are a lot of nasty issues
[15:39:38] <milimetric>	 but the advantage to having a real db would be pretty great
[15:40:05] <milimetric>	 i meet a lot of people at the foundation that say they hate SQL, I think they really mean they hate mysql
[15:40:32] <YuviPanda>	 milimetric: indeed.
[15:40:39] <YuviPanda>	 milimetric: it's a lot of work tho :(
[15:40:53] <milimetric>	 yea, no doubt
[15:40:55] <grrrit-wm>	 (PS1) Erik Zachte: minus negwikibooks, plus zhwikivoyage [analytics/wikistats] - https://gerrit.wikimedia.org/r/126442
[15:42:50] <grrrit-wm>	 (PS1) Erik Zachte: dataset2 -> dataset 1001 [analytics/wikistats] - https://gerrit.wikimedia.org/r/126468
[15:44:16] <grrrit-wm>	 (PS1) Erik Zachte: updated binaries and html index file [analytics/wikistats] - https://gerrit.wikimedia.org/r/126487
[15:45:45] <YuviPanda>	 milimetric: we're hiring some more android folks, so I will be doing some analytics work once the app releases (a month?). Will be building a non-limn dashboard, most probably using vega :) Do let me know if you've thoughts
[15:46:44] <milimetric>	 cool YuviPanda, our vega project might be just getting started around the same time
[15:46:54] <milimetric>	 right now the designers are trying to figure out how to lay out the dashboard
[15:47:01] <milimetric>	 we should definitely keep in touch
[15:47:06] <YuviPanda>	 will do, milimetric
[15:48:35] <grrrit-wm>	 (PS1) Erik Zachte: new absolute monthly counts, plus added dates to ignore to blacklist [analytics/wikistats] - https://gerrit.wikimedia.org/r/126546
[16:14:46] <grrrit-wm>	 (CR) Hashar: "Sorry you would need a patchset. recheck does not rerun the tests until we get them to run in isolated sandboxes :-(" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[16:25:36] <grrrit-wm>	 (PS1) Erik Zachte: new images for wikistats portal [analytics/wikistats] - https://gerrit.wikimedia.org/r/126665
[16:30:17] <Ironholds>	 YuviPanda, what's "the UA RfC"?
[16:30:33] <YuviPanda>	 Ironholds: qchris pointed it out to me in an email in that thread. let me find it
[16:30:49] * qchris  denies everything

[16:31:46] <qchris>	 http://tools.ietf.org/html/rfc2616
[16:32:02] <qchris>	 YuviPanda ^
[16:33:01] <YuviPanda>	 Ironholds: http://tools.ietf.org/html/rfc2616#section-3.8 and the section just before http://tools.ietf.org/html/rfc2616#section-3
[16:33:01] <YuviPanda>	 so the UA mentions in the HTTP 1.1 RfC
[16:33:02] <YuviPanda>	 Ironholds: let me put those in gerrit
[16:33:33] <YuviPanda>	 Ironholds: http://lists.wikimedia.org/pipermail/analytics/2014-March/001730.html
[16:33:47] <Ironholds>	 yes, your user agent is not compliant with that
[16:33:56] <Ironholds>	 if it was, ua-parser would parse it
[16:34:02] <Ironholds>	 I know this because I just altered it TO comply, and it parses ;p
[16:34:20] <YuviPanda>	 aha! what's the fix?
[16:34:28] <Ironholds>	 semicolon after the OS version number.
[16:34:32] <Ironholds>	 As qchris's email has
[16:34:45] <Ironholds>	 I'll leave a note
[16:34:52] <YuviPanda>	 Ironholds: ah, indeed. Like https://gerrit.wikimedia.org/r/#/c/126247/ does
[16:35:18] <Ironholds>	 then why did the example UAs in the email not include em?
[16:35:37] <YuviPanda>	 Ironholds: it's just not been implemented in the iOS version, and I added it to the Android version after Adam sent that email
[16:35:43] <YuviPanda>	 Ironholds: hence the confusion.
[16:35:52] <Ironholds>	 ahh
[16:36:08] <YuviPanda>	 I should've replied to that email, but it was late when I saw i so didn't.
[16:36:12] <YuviPanda>	 and yay for confusion resolution :)
[16:37:21] <YuviPanda>	 Ironholds: Can you comment there or should I?
[16:37:27] <Ironholds>	 I'm gonna
[16:37:34] <Ironholds>	 I'm just testing/resolving the iPhone one
[16:37:48] <Ironholds>	 so far I've got the modification necessary to make it recognise it's an iPhone. From there..
[16:38:24] <YuviPanda>	 ok!
[16:38:41] <YuviPanda>	 and woot for doing these *before* the app releases :)
[16:39:15] <Ironholds>	 well, I said I'd test em
[16:39:27] <Ironholds>	 and I don't particularly want to be on the hock for parsing the unparseable, or writing weird rules we have to maintain
[16:39:54] <YuviPanda>	 indeed, but if they do match the RfC they should parse.
[16:40:15] <Ironholds>	 except they didn't as provided to me!:P
[16:40:21] * Ironholds  checks the iOS patch

[16:40:34] <YuviPanda>	 indeed.
[16:40:41] <YuviPanda>	 Ironholds: incomplete they are.
[16:41:09] <YuviPanda>	 the android patch should be merged shortly. then they will be WikipediaApp/v2.0 (Android/4.4;tablet)
[16:41:13] <YuviPanda>	 err, Tablet
[16:41:17] <YuviPanda>	 should there be a space between ; and tablet?
[16:42:41] <Ironholds>	 no slash!
[16:42:45] <Ironholds>	 Android 4.4;
[16:43:07] <Ironholds>	 and, good question
[16:43:45] <Ironholds>	 answer comes back 'it's irrelevant, ua-parse doesn't have device class parsing yet, just device parsing'
[16:43:52] <Ironholds>	 so, if you can provide the actual device, that'd be grand
[16:44:02] <Ironholds>	 if it's just tablet versus mobile it probably won't pick it up
[16:44:02] <YuviPanda>	 actual device as in?
[16:44:09] <YuviPanda>	 HTC Explorer vs Nexus 4?
[16:44:14] <Ironholds>	 HTC ([A-Z][a-z0-9]+) Build'
[16:44:17] <YuviPanda>	 we explicitly didn't put that in because of privacy concerns.
[16:44:29] * Ironholds  sighs

[16:44:32] <Ironholds>	 okay. hurgh.
[16:44:45] <grrrit-wm>	 (CR) QChris: "> Sorry you would need a patchset. recheck does not rerun the tests" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[16:44:48] <Ironholds>	 then include tablet however the hell you want
[16:44:56] <Ironholds>	 because either I'll have to write a regex for it or we won't use it ;p
[16:45:06] <YuviPanda>	 heh
[16:45:06] <Ironholds>	 but either way format doesn't matter as long as it doesn't interfere with OS identification
[16:45:11] <YuviPanda>	 so
[16:45:21] <YuviPanda>	 WikipediaApp/v2.0 (Android 4.4; Tablet)
[16:45:21] <Ironholds>	 WikipediaApp/v2.0 (Android 4.4; tablet)
[16:45:24] <Ironholds>	 hah!
[16:45:24] <YuviPanda>	 hah!
[16:45:26] <Ironholds>	 ...
[16:45:27] <YuviPanda>	 hah!
[16:45:59] <YuviPanda>	 missed that. my mind reader mustn't be working
[16:46:23] <Ironholds>	 user agents: reducing confidence in the theory that we are different people since 1999
[16:46:27] <Ironholds>	 permission to post this to bash?
[16:47:18] <YuviPanda>	 hehe
[16:47:24] <YuviPanda>	 definitely
[16:47:42] <YuviPanda>	 Ironholds: I updated the Android patch to provide that.
[16:48:06] <Ironholds>	 grand!
[16:50:01] <Ironholds>	 alright, patch note added
[16:50:26] <Ironholds>	 y'know, ua-parser is implemented in python
[16:50:30] <Ironholds>	 so testing is trivial
[16:50:34] <YuviPanda>	 Indeed
[16:50:54] <YuviPanda>	 now need to figure out who to hassle for my stat1003 access (I responded on the thread)
[16:50:59] <YuviPanda>	 no ottomata :(
[16:51:13] <Ironholds>	 so are we not including device for iOS devices, where it's just iPhone versus iPod versus iPad?
[16:52:26] <YuviPanda>	 good question. depends on what makes more sense to you.
[16:52:32] <YuviPanda>	 but there's also iPod touch
[16:52:38] <Ironholds>	 yep
[16:52:42] <Ironholds>	 I more meant from a privacy POV
[16:52:47] <YuviPanda>	 but yeah, we can just make it iPhone vs iPad and let it be, or add the ; tablet
[16:52:53] <YuviPanda>	 current idea is to mimic android
[16:52:54] <Ironholds>	 point being: fewer devices, so that's good, buuut
[16:53:05] <YuviPanda>	 I mean, we won't be reporting iPhone 5s, for example.l
[16:53:20] <YuviPanda>	 s/l$//
[17:02:35] <Ironholds>	 yup
[17:06:42] <Ironholds>	 YuviPanda, got it to work
[17:06:52] <YuviPanda>	 Ironholds: woot!
[17:06:58] <YuviPanda>	 Ironholds: what's the form it worked in?
[17:08:57] <Ironholds>	 YuviPanda, see email and the patch
[17:09:05] <YuviPanda>	 Ironholds: woot. awesome!
[17:09:15] <YuviPanda>	 Ironholds: I think adam will fix iOS, and android patch should be merged shortly
[17:09:18] <Ironholds>	 cool
[17:09:20] <YuviPanda>	 Ironholds: ty!
[17:09:23] <Ironholds>	 I can always patch it myself, but eeeeh.
[17:09:24] <Ironholds>	 cool
[17:14:01] <YuviPanda>	 hey ottomata!
[17:14:19] <ottomata>	 yoyooo
[17:14:20] <YuviPanda>	 I remember responding to you about stat1 access moving over, but looks like I don't have stat1003 access? nor bast1001 :(
[17:14:36] <ottomata>	 hmmmmMMmm
[17:15:16] <ottomata>	 ! i don't know how that slipped by Yuvi!
[17:15:29] <ottomata>	 fixing...
[17:15:48] <YuviPanda>	 ty ottomata!
[17:19:30] <halfak>	 Hey ottomata.  Is it possible that some cron jobs are still running from old stat1?
[17:20:33] <ottomata>	 yes!  i didn't disable any of them, i just moved the new ones over
[17:20:39] <ottomata>	 so they are probably running in both places?
[17:21:09] <halfak>	 There are some problematic queries running against s1 and I suspect they are attached to old, defunct crons.
[17:21:58] <halfak>	 Sadly, they are using the common "research" user for connecting to MySQL.
[17:22:15] <halfak>	 Do you know how I might track down the source?
[17:23:49] <ottomata>	 well, afaik all crons are still running on stat1
[17:24:02] <ottomata>	 stat1 is still up, log in and see if any of your crons are there, and if you don't want them, just disable them?
[17:24:24] <Ironholds>	 ahh
[17:24:26] <halfak>	 Oh yeah.  Totally could do that for myself.  This issue isn't one of my crons though.
[17:24:33] <Ironholds>	 ottomata, related question, what happens to crons of people who leave?
[17:25:02] <ottomata>	 if we remove their account...i'm not actually sure!  I assume they won't run...but they might
[17:25:11] <ottomata>	 i think we don't actualy remove accounts though
[17:25:15] <ottomata>	 usually we just disable ssh keys
[17:25:33] <halfak>	 Could we stop all the crons of people with disabled keys?
[17:26:06] <Ironholds>	 Only we're pretty sure some of erosen's scripts are still triggering (or were)
[17:26:14] <halfak>	 DarTar: ^
[17:26:51] <ottomata>	 let's see...
[17:27:07] <DarTar>	 hey
[17:27:41] <DarTar>	 so I am not sure if erosen’s scripts are reused by other people, aren’t they generating data for the geodashboards used by grantmaking?
[17:28:01] <ottomata>	 halfak: i do see that you ahve crons on stat1
[17:28:10] <ottomata>	 that are probably still running
[17:28:10] <ottomata>	 including some mysql stuff
[17:28:14] <halfak>	 I just killed 'em ottomata
[17:28:18] <halfak>	 like 2 secs ago
[17:28:26] <DarTar>	 ottomata: I do have crons, I can port them
[17:28:54] <halfak>	 DarTar, should have been already ported for you.
[17:29:01] <DarTar>	 ok cool
[17:29:10] <ottomata>	 DarTar, they have all been ported ja
[17:29:13] <DarTar>	 danke
[17:29:14] <ottomata>	 but I did not remove them from stat1
[17:29:25] <halfak>	 DarTar, if erosen's account is running old crons that we need, we need to take them over.
[17:29:29] <ottomata>	 so they are probably running on stat1 and stat1003 right now
[17:29:35] <ottomata>	 i don't see erosen crons on stat1
[17:29:36] <halfak>	 I'll add a card to audit old crons.
[17:29:40] <DarTar>	 if the rsync to stat1001 is in place from stat1003 I can safely disable them all from stat1
[17:29:41] <ottomata>	 checking other hosts
[17:29:52] <ottomata>	 DarTar it should be
[17:30:05] <ottomata>	 also i think the rsync from stat1 won't work anymore as of yesterday anyway
[17:30:06] <DarTar>	 ok
[17:30:10] <ottomata>	 hmm, or today
[17:30:13] <ottomata>	 sorry, yeah this morning
[17:30:16] <DarTar>	 alright
[17:33:10] <DarTar>	 hm ottomata what’s stat1003 full hostname?
[17:33:17] <ottomata>	 stat1003.wikimedia.org
[17:33:30] <DarTar>	 I can’t SSH into it for some reason
[17:34:02] <DarTar>	 my connection hangs
[17:34:03] <ottomata>	 you have to go through bast1001
[17:34:08] <ottomata>	 are you?
[17:34:09] <DarTar>	 ah, right
[17:34:12] <halfak>	 Firewalls!
[17:34:12] <DarTar>	 makes sense
[17:35:38] <Ironholds>	 heh
[17:35:45] <DarTar>	 dartar@bast1001:~$ ssh stat1003.wikimedia.org
[17:35:45] <DarTar>	 Permission denied (publickey).
[17:35:52] <DarTar>	 huh
[17:35:58] <ottomata>	 DarTar
[17:36:03] <ottomata>	 you can't ssh directly without forwarding your ssh key
[17:36:15] <ottomata>	 https://wikitech.wikimedia.org/wiki/Server_access_responsibilities#SSH
[17:36:30] <DarTar>	 heh, I forgot that
[17:36:36] <ottomata>	 you should use ProxyCommand, though, not forward
[17:36:37] <ottomata>	 bu tja
[17:38:09] <DarTar>	 I thought my .ssh/config was already set up touse ProxyCommand, I’ll have to check again
[18:37:39] <ori>	 halfak, Ironholds, DarTar: the hangout link on the calendar item for the research showcase needs to be updated
[18:38:08] <DarTar>	 yes, I’ll remove it, sorry for the confusion
[18:39:07] <DarTar>	 wow, hangout doesn’t allow me to remove a video link once it’s added
[18:39:08] <DarTar>	 clever
[18:39:13] <DarTar>	 I’ll ask office it
[18:39:41] <ori>	 i don't have the right link
[18:39:46] <ori>	 is there no showcase at the moment?
[18:41:30] <ori>	 ^ DarTar
[18:42:06] <DarTar>	 ori: https://www.youtube.com/watch?v=Pps__TkfrMs
[18:42:27] <ori>	 thank you
[18:42:29] <DarTar>	 and #wikimedia-research for discussion
[18:42:47] <ori>	 moar channels
[18:46:46] <grrrit-wm>	 (CR) Hashar: "@qchris and I upgraded Zuul today, I should be able to have recheck to retrigger test jobs for trusted folks :-]" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[19:09:16] <DarTar>	 ori: that channel was created around 2005-6 :p
[19:09:28] <DarTar>	 long before #wikimedia-analytics existed
[19:12:54] <YuviPanda>	 is there a doc setup on how to reach the mysql slaves from stat1003? I've forgotten...
[19:16:36] <ottomata>	 YuviPanda: maybe this? https://wikitech.wikimedia.org/wiki/Analytics/Data_access#Analytics_slaves
[19:18:45] <YuviPanda>	 ottomata: is what I need. ty!
[19:27:29] <Ironholds>	 yay!
[19:27:33] <Ironholds>	 people are using my documentation
[19:27:48] <ottomata>	 thanks for that Ironholds :)
[19:27:54] <ottomata>	 I will try to keep it up to date too, as things change
[19:27:57] <ottomata>	 like, more data going into hive
[19:28:02] <ottomata>	 etc.
[19:28:17] <Ironholds>	 no problemo!
[19:28:20] <Ironholds>	 congrats on bits, btw :)
[19:30:49] <YuviPanda>	 ottomata: is there a way I can see a stream of all events that are coming in? Mine might not be validating, and I need to verify that
[19:34:28] <YuviPanda>	 milimetric: ^
[19:34:39] <YuviPanda>	 or more to the point, how do I debug if my events are making it at all vs if they aren't validating?
[19:34:40] <ottomata>	 events like event log?
[19:34:46] <YuviPanda>	 ottomata: EventLogging, yeah.
[19:34:57] <ottomata>	 hm, I think you need to have an account on vanadium for that, not sure,
[19:35:00] <YuviPanda>	 ottomata: https://wikitech.wikimedia.org/wiki/EventLogging#Varnish mentions vanadium
[19:35:08] <ottomata>	 afaik, eventlogging on stat1003 is just archived logs
[19:35:09] <ottomata>	 hourly i think
[19:35:10] <YuviPanda>	 right
[19:35:13] <YuviPanda>	 oh wat
[19:35:20] <YuviPanda>	 no, the mysql replication is instantaneous, I think?
[19:35:42] <ottomata>	 oh in the db?
[19:35:43] <ottomata>	 i guess so
[19:35:48] <ottomata>	 i'm thinking on the disk
[19:35:49] <ottomata>	 https://wikitech.wikimedia.org/wiki/EventLogging#Check_if_the_data_I_am_logging_is_coming_through_properly.3F
[19:35:51] <ottomata>	 is that helpful?
[19:36:20] <YuviPanda>	 ottomata: no because zsub isn't installed
[19:36:24] <YuviPanda>	 on stat1003
[19:36:25] <ottomata>	 oh hm
[19:36:27] <ottomata>	 hmmmm
[19:36:38] <milimetric>	 oh yeah, we must've missed that in the migration
[19:36:45] <milimetric>	 as in - it probably wasn't puppetized
[19:36:54] <YuviPanda>	 ah, right.
[19:36:59] <YuviPanda>	 should I make a patch?
[19:37:07] <milimetric>	 but that's where i'd check the zeromq stream usually
[19:37:30] <milimetric>	 sure YuviPanda, patches welcome, but I think after that we'd need to subscribe it as well, I never did that but I guess I could learn
[19:37:41] <milimetric>	 also, YuviPanda, any problem with just checking the db?
[19:37:54] <YuviPanda>	 milimetric: yeah, it isn't there in the db
[19:38:03] <YuviPanda>	 milimetric: so I need to get the raw logs which will tell me if it is being dropped
[19:38:30] <milimetric>	 ok, makes sense YuviPanda, yeah, events might not be validating or there could be other problems between the stream and its being written to the db
[19:38:30] <ottomata>	 i can patch it for install
[19:38:32] <ottomata>	 few secs
[19:38:39] <milimetric>	 so yeah, 1. install 2. subscribe
[19:38:39] <YuviPanda>	 ottomata: woo!
[19:39:04] <YuviPanda>	 milimetric: yeah, subscribe should be simple if there are no firewall issues
[19:48:40] <grrrit-wm>	 (CR) QChris: "recheck" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[19:50:22] <hashar>	 qchris: na it is not working yet :-D
[19:50:28] <qchris>	 :-D
[19:50:34] <qchris>	 Was trying ... just in case.
[19:50:41] <hashar>	 we never know
[19:50:45] <qchris>	 But I am replying to your comment.
[19:51:09] <qchris>	 I learnt that hashar is quick ... so I wanted to check before requesting it :-P
[19:51:19] * hashar  blushes

[19:51:36] <hashar>	 with the old zuul version, it would retrigger based solely on the comment
[19:51:53] <hashar>	 with the new version, we can combine requirements
[19:52:15] <grrrit-wm>	 (CR) QChris: "> I upgraded Zuul today" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[19:52:16] <hashar>	 so we could get  "comment is 'recheck'" AND "email is whitelisted"
[19:52:26] <qchris>	 Mhmmmm.
[19:52:30] <qchris>	 Sounds supergood.
[19:53:03] <qchris>	 Would it be hard to get that working?
[19:53:15] * hashar  ponders

[19:53:29] <hashar>	 it is 10pm and I havent finished my cognac + I need some sleep
[19:53:40] <qchris>	 Hahaha. Cognac first.
[19:53:45] <hashar>	 on a serious note, no that is not too hard
[19:53:56] <hashar>	 just have to read the doc, figure out what is needed, test it locally and push!
[19:54:32] <qchris>	 It would help me a lot :-)
[19:59:37] <hashar>	 the hardest part is figuring out whether a bug got filled
[20:03:42] <YuviPanda>	 milimetric: fwiw, I can use zsub as mentioned in the wikitech EL Page to debug, so all is well
[20:04:01] <milimetric>	 ok, awesome YuviPanda
[20:07:05] <grrrit-wm>	 (CR) Hashar: "qchris : the zuul conf would be https://gerrit.wikimedia.org/r/#/c/126838/ :-D" [analytics/wikistats] - https://gerrit.wikimedia.org/r/126264 (owner: Erik Zachte)
[20:07:35] <qchris>	 hashar: That's not fair.
[20:07:41] <qchris>	 I am not done with filing the bug ...
[20:07:47] <hashar>	 well once you did
[20:07:49] <qchris>	 Meh... I knew you were qucik :-)
[20:07:54] <hashar>	 you can amend that change and Bug: 123456
[20:08:08] <hashar>	 then I can redirect the necessary bike shed to that bug report :-]
[20:08:12] <qchris>	 And I can make the bug's description "merge change 126838"
[20:08:24] <hashar>	 to be honest, I already thought about that change during the dinner
[20:08:31] <qchris>	 Cheater!
[20:09:09] <hashar>	 before having a kid, I was dreaming about problems, found a solution during the shower.  Would then show up at 9:30 exactly, code, and by 10am my day was over :]
[20:09:25] <qchris>	 Hahaha.
[20:10:07] <hashar>	 the recheck thing, I will got to test it out on labs
[20:10:31] <hashar>	 I have to rebuild Zuul/Jenkins in labs for dev purposes tomorrow afternoon. The change above is going to  an excellent motivation.
[20:22:08] <YuviPanda>	 hmm, any idea  why making a request with 'https://bits.wikimedia.org/event.gif?%7B%22schema%22%3A%22MobileWikiAppEdit%22%2C%22revision%22%3A%228174633%22%2C%22event%22%3A%7B%22action%22%3A%22start%22%2C%22userName%22%3A%22yuvitest44%22%2C%22editSessionToken%22%3A%22cf724bdf-e4d7-4b2f-9db0-0b5a3e129ba7%22%7D%7D' isn't even doing anything?
[20:22:13] <YuviPanda>	 nothing on the varnishcsa logs either
[20:22:46] <YuviPanda>	 milimetric: ^
[20:22:54] <YuviPanda>	 (or whoever I should poke about EL)
[20:23:17] <YuviPanda>	 ah, non 200 code
[20:23:19] * YuviPanda  investigates

[20:25:02] <YuviPanda>	 oh, nevermind. it is returning a 204 as it should
[20:25:05] <YuviPanda>	 but nothing on the server
[20:25:06] <YuviPanda>	 HWY
[20:25:07] <YuviPanda>	 WHY
[20:26:39] <YuviPanda>	 ah cool
[20:26:43] <YuviPanda>	 it does hit varnishcsa
[20:26:45] <YuviPanda>	 ignore me, sorry
[20:46:34] <YuviPanda>	 ok, so it hits varnish but somehow gets dropped in the middle and I've no idea why :(
[20:49:01] <YuviPanda>	 {"schema":"MobileWikiAppEdit","revision":8188113,"event":{"action":"start","userName":"yuvitest44","editSessionToken":"77078656-8520-4c5d-832a-7b37b4ba7cf8"}} is the payload
[20:49:09] <YuviPanda>	 and it is hitting https://bits.wikimedia.org/event.gif
[20:49:14] <YuviPanda>	 and  I get back a 204
[20:49:35] <YuviPanda>	 I looked at both EventCapsule and event itself and they both seem to validate. Unsure why it isn't coming through
[20:50:31] <YuviPanda>	 anyone?
[20:52:26] * YuviPanda  pokes milimetric and halfak randomly

[20:52:52] <halfak>	 In meeting -- receptive to random poke in ~ 45 min
[20:53:01] <YuviPanda>	 halfak: oh, I guess that's all of you then.
[20:53:07] <YuviPanda>	 halfak: i'll just wait. Thanks for that!
[21:13:46] <milimetric>	 hm, YuviPanda, I don't know how to debug that, could you file a bug and cc me on it?  I'll get it prioritized as we have to learn this sooner rather than later
[21:13:55] <YuviPanda>	 milimetric: alright
[21:15:04] <milimetric>	 and ori, please don't work on this bug that YuviPanda's about to file.  I know you could probably fix it in half a second, but we gots to learn :)
[21:25:19] <YuviPanda>	 milimetric: https://bugzilla.wikimedia.org/show_bug.cgi?id=64024
[21:26:03] <YuviPanda>	 milimetric: hopefully cc'd you in the right mail
[21:26:24] <YuviPanda>	 milimetric: I'll be off to sleep in a bit
[21:27:02] <milimetric>	 yep YuviPanda, thanks and that includes me
[21:27:27] <YuviPanda>	 milimetric: :D
[21:27:41] <YuviPanda>	 milimetric: yw! And do note that we are hoping to ship the entire thing in a couple of weeks
[21:28:27] <milimetric>	 yep YuviPanda, this is getting high importance
[21:28:34] <YuviPanda>	 milimetric: ty!
[21:28:52] <YuviPanda>	 milimetric: it might be me doing something stupid (probable!), fwiw :)
[22:04:36] <Ironholds>	 YuviPanda, I don't read code.
[22:04:53] <Ironholds>	 Could you take a look at the new iOS patch and tell me if it incorporates an underscore between version numbers? (I.e. 4_4)
[22:04:57] <Ironholds>	 I can't see any way it could, but.
[22:11:25] <Ironholds>	 wait, nm
[22:13:08] <dr0ptp4kt>	 Ironholds, the ios doesn't try to make an underscore of it
[22:13:18] <dr0ptp4kt>	 do you want that? or are dots okay?
[22:13:32] <dr0ptp4kt>	 Ironholds: you should hang out on #wikimedia-mobile, too!
[22:14:12] <dr0ptp4kt>	 Ironholds: apple requires the build version to be in #.#.#, so it's easy enough to do a find and replace if an underscore is needed. i just rebased, by the way, didn't make any other changes
[22:17:26] <Ironholds>	 dr0ptp4kt, yeah, just tested it
[22:17:29] <Ironholds>	 it works fine :)
[22:17:33] <Ironholds>	 hence my +1
[22:17:45] <dr0ptp4kt>	 Ironholds: cool
[22:17:46] <Ironholds>	 (sorry, popped out for a smoke)
[22:18:25] <Ironholds>	 argh, rebase!
[22:18:27] * Ironholds  +1s again

[22:18:36] <Ironholds>	 we reaaally need that gerrit config option set
[23:15:32] <halfak>	 Hey YuviPanda.  Still around?
[23:16:08] <halfak>	 Sorry for the delay.  :\