[07:01:25] New patchset: Rfaulk; "fix. self refs in class methods." [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72687 [07:01:25] New patchset: Rfaulk; "mod/fix. raise exceptions on abstract class method calls/pep8" [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72688 [07:01:25] New patchset: Rfaulk; "add. FileBroker definition." [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72689 [07:01:56] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72687 [07:02:13] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72688 [07:02:29] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72689 [08:39:17] New patchset: Rfaulk; "add. remove functionality." [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72694 [08:39:17] New patchset: Rfaulk; "add. update functionality." [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72695 [08:39:18] New patchset: Rfaulk; "add. get functionality." [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72696 [08:39:43] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72694 [08:39:57] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72695 [08:40:13] Change merged: Rfaulk; [analytics/user-metrics] (repair_runtime) - https://gerrit.wikimedia.org/r/72696 [13:23:50] mmooooooorning!! [13:25:54] ottomata can you handle https://rt.wikimedia.org/Ticket/Display.html?id=5403 ? [13:27:06] drdee: morning :) [13:28:21] hye average! [13:48:06] yo ottomata [13:48:10] yoyo qchris [13:48:19] ottomata can you handle https://rt.wikimedia.org/Ticket/Display.html?id=5403 ? [13:48:21] hi drdee [13:48:27] then qchris will have access to kraken [13:48:39] Yippie! [13:48:40] tnegrin has approved the request and it has been in the queue for more than 3 days [13:49:43] yooy [13:49:43] can do [13:51:15] qchris what is your shell username? [13:51:21] ottomata: qchris [13:51:28] hm [13:51:33] drdee: do we have showcase this week ? [13:51:55] qchris, I can't find you in admins.pp [13:52:16] oh yeah, wednesday, seems to be in calendar [13:52:18] ottomata: I doubt I'm in there. I am no admin. [13:52:37] right but that's where users are [13:52:38] um [13:52:45] do you have a shell account on a production box yet? [13:53:06] ottomata: no. [13:53:09] ohhhh [13:53:10] ok [13:53:12] Just labs. [13:53:19] ohoo k [13:53:21] hmmm, ok [13:53:27] what username do you use to log into labsconsole? [13:53:34] also, do you have access to office.wikimedia.org? [13:54:07] Not quite sure what yu mean by labsconsole ... qchris is my shell name on labs. QChris in my username. [13:54:17] No, I do not have access to office.wikimedia.org. [13:54:19] hmmmMmmm [13:54:32] do you have a wikitech login? [13:54:49] ottomata: yes, QChris is my wikitech login [13:54:59] ok, pretty sure that is your ldap name, hmm, ok that's good [13:55:13] Yes, that's my LDAP name. [13:55:21] ok, then, on wikitech, can you log in and create a page under your username and add your public ssh key there? [13:55:37] the ops guys want me to accept your ssh key only through services that you can authenticate yourself with [13:55:39] wikitech will do [13:56:16] ok. [13:58:28] ottomata: https://wikitech.wikimedia.org/wiki/User:QChris/Ssh [14:00:04] dankje [14:13:13] ok qchris, i manually ran pupet on some relevant nodes [14:13:45] ottomata: Yay. Thanks. [14:14:00] * qchris now has superpowers :-) [14:14:26] hey ottomata, when I try to ssh into limn0, I get: "ssh_exchange_identification: Connection closed by remote host" [14:14:28] you get the same? [14:14:38] qchris, you should read this doc: [14:14:38] https://wikitech.wikimedia.org/wiki/Server_access_responsibilities [14:14:42] (I gotta run out for 15 min., brb) [14:14:51] especially the bit about ssh ProxyCommand stuff [14:15:01] ottomata: Ok, thanks. [14:15:30] milimetric: i can log into labs fine [14:15:36] but I use a different bastion host than you I think [14:54:30] drdee, if you think the reportcard looks good, I can publish it to prod, I put it in sign off for now [14:54:46] great! what's the link? [15:23:07] http://test-reportcard.wmflabs.org/ drdee [15:23:10] same as always [15:23:25] sorry :( [15:54:55] ok, i'm going to grab some lunch before standup today [15:54:56] back in a bit [16:03:14] New patchset: Erosen; "fully separates celery tasks from sqlalchemy mapped class" [analytics/wikimetrics] (master) - https://gerrit.wikimedia.org/r/72739 [16:03:29] Change merged: Erosen; [analytics/wikimetrics] (master) - https://gerrit.wikimedia.org/r/72739 [16:04:10] milimetric: tests are passing and jobs are running, just haven't totally figured out job persistence scheme [16:05:21] woa! [16:05:22] nice man [16:35:31] erosen, I pushed the re-enabled queue in tests' __init__ [16:35:37] but yeah, everything looks good [16:35:39] nice [16:35:41] still reading over your changes [16:36:12] oh wait you did that on purpose? [16:36:21] so you're thinking we'd always run celery externally? [16:36:34] not necessarily [16:36:40] though it is essential for debugging [16:36:47] until I figure out how to get celery logs [16:37:05] i think I might create a central task_logger in configurables [16:37:12] milimetric: I also created a PersistentJob class, but I haven't fleshed out the interaction between Job and PersistenJob [16:38:08] cool [16:38:12] i like the central task logger [16:38:16] and logging to files is totally fine [16:38:21] I mean it'll have to do that in prod anyway [16:38:43] just make it configurable [16:39:18] but yeah, so basically as long as the classes aren't SQLAlchemy classes and they have that filter=task_method thing, they work huh? [16:39:50] basically [16:39:55] it was mostly a matter of removing code [16:40:00] there is a small bug in celery i think [16:40:09] i'll ask you about it at some point [16:40:37] well, dude, awesome job fighting through this [16:40:56] i think this was one of the last, if not the last weird spider web thing [16:41:03] now it should be a clean foundation :) [16:41:04] it was really relatively straightfoward once we decided how to deal with it yesterday [16:41:17] i certainly hope so [16:52:58] milimetric: can you point the way to commit squashing / gerrit rebasing? [16:53:16] maybe [16:53:19] what's the situation [16:53:45] I commitd something, pulled your changes, made another commit [16:53:51] and now I want to review [16:53:54] New patchset: Stefan.petrea; "Debianization to conform with Ops requirements" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [16:54:11] and when you do git review it cries right? [16:54:17] I'm currently 4 comits ahead of origin/master [16:54:21] yup [16:54:30] easiest way: just push [16:54:33] :) [16:54:38] hehe [16:54:42] it hink that maybe the right thing here [16:54:43] k [16:54:51] because you'd have to squash my history in there (my fault for not git-reviewing) [16:55:02] np [16:55:24] milimetric: k, done [16:55:25] so yeah, squashing my history is "a bad thing" because it's shared [16:55:25] easy [16:55:36] good point [16:55:59] but shouldn't I be able to rebase at least? [16:56:22] but normally if you were working with nice partners that always git-reviewed, you could just git commit --amend [16:56:41] interesting, I've never fully understood --amend [16:57:16] yeah, I think it changes the last commit [16:57:47] I guess I've only encountered it for changing the message [16:57:51] yep [16:57:57] which is what it Should be used for [16:57:57] but you mean that I could just add new changes to an old commit? [16:58:11] but gerrit abuses it for adding a new patchset [16:58:17] yea [16:58:22] i see [16:59:58] 5seconds [17:29:38] milimetric: quick q on the jobs/create interface [17:30:11] when I click run job with no cohort selected, it pops up a nasty alert box with unrendered html [17:30:37] i suspect this is a simple fix, but I think we should maybe deal with that as part of 696 [17:32:00] * average just rolled .. another cigarette [17:35:05] New patchset: Stefan.petrea; "Debianization to conform with Ops requirements" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [17:39:41] milimetric: ^^ [17:40:02] milimetric: we don't have to deal with this now, but we should find some time today [17:40:18] sorry erosen, I'm talking to sucheta on a hangout [17:40:24] I'll be back shortly [17:40:26] milimetric: no worries [17:40:30] but yes, today, soon [17:40:40] New patchset: Stefan.petrea; "Debianization to conform with Ops requirements" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [17:41:08] ok, LINTIAN Error free libdclass [17:41:14] (some warnings still present) [18:19:49] New patchset: Stefan.petrea; "Debianization to conform with Ops requirements" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [18:20:04] Hallo. Is there any way to see more languages in http://reportcard.wmflabs.org/graphs/articles_per_day ? [18:23:14] drdee: ^ maybe you'll know [18:24:16] hi aharoni [18:24:24] hi milimetric [18:24:31] so the source for that graph comes from wikistats [18:24:41] and all the languages we track are available there [18:25:07] probably changing limn-data/datasources/rc_new_article_count.yaml to add more columns also ? [18:25:10] milimetric: ^^ [18:25:22] for the languages I mean [18:25:29] well, it's not that simple [18:25:35] because the wikistats data is in a custom csv format [18:25:41] and I'd have to parse the new columns out [18:26:08] ok [18:26:32] but average, maybe you could show aharoni where the "articles per day" analysis is in wikistats, I am very bad at finding my way around in there [18:27:03] I can find the tables in stats.wikimiedia.org . [18:27:04] milimetric: what's the name of the CSV you use from Wikistats ? [18:27:22] I was wondering whether it's possible to see the pretty graphs for more languages and not just for the big five. [18:27:36] You know, to easily point people there. [18:27:55] Most people don't have the patience for the tables at stats.wikimiedia.org . [18:28:00] it's a custom csv average, erik zachte sends it's wikilytics_in_wikistats_core_metrics.csv [18:28:29] The Hebrew Wikipedia is celebrating its tenth anniversary today, and I'm being interviewed by radio and TV stations about it, and I want to be prepared in case they ask about numbers ;) [18:28:32] yeah aharoni, I'll ask to see if we can do that [18:28:44] cool, thanks. [18:28:59] hey that's very cool about Hebrew Wikipedia [18:41:50] hey erosen, ok [18:41:53] yo yo [18:41:55] that nasty html popup [18:42:05] yeah [18:42:07] i did that on purpose [18:42:08] want to hangout? [18:42:09] k [18:42:10] sure [18:42:19] just cause I have other qs [18:47:02] brb [20:24:27] milimetric: back, btw. I'm gonna mess with bytes added a bit and then start on the job status stuff [20:24:38] let me know if you want to hangout\ [20:24:49] what are the warnings average? [20:25:16] erosen: job status: you mean progress? [20:26:13] drdee: we're finishing up the back end work associated with creating new request [20:26:30] k [20:26:30] it sort of bleeds into job management though, which involves checking status [20:39:06] (sorry, wireless problems, trying a different adapter) [21:15:47] milimetric: did you ever figoure out how to do server side wtform validation [21:15:57] yeah, it's doing it now [21:15:59] with metrics [21:16:21] if you look at the bytes_added test, it first does a basic "does it validate" check [21:16:30] (the second and third tests, not the first one) [21:16:35] great [21:16:37] i'll look there [21:51:50] drdee: http://stats.wikimedia.org/wikimedia/squids/TablesPageViewsMonthlySquidsMobile.htm [21:51:57] drdee: fresh out of the cron job [21:52:07] Erik's cron job [22:07:06] New patchset: Milimetric; "starting to port the cohort upload" [analytics/wikimetrics] (master) - https://gerrit.wikimedia.org/r/72847 [22:07:20] Change merged: Milimetric; [analytics/wikimetrics] (master) - https://gerrit.wikimedia.org/r/72847 [22:07:31] hey erosen, I finished some basic cohort upload porting [22:07:34] haven't tested it at all [22:07:36] nice [22:07:41] just made sure all the old tests still pass [22:07:46] good start [22:07:50] what is your plan for the evening? [22:07:52] we promised we'd get the other stuff past review though [22:07:55] so that's my plan :) [22:08:00] wanna hang out and coordinate? [22:08:04] y [22:08:05] a [22:08:42] milimetric: diederik and toby are in regular hangout [22:08:45] yep [22:08:50] i'll make