[04:55:51] Change merged: Yuvipanda; [analytics/limn-mobile-data] (master) - https://gerrit.wikimedia.org/r/60610 [04:58:31] YuviPanda: I haven't forgotten about the character encoding bug you filed, by the way. I'm just behind, as usual. [04:58:41] :D [04:58:59] it'll be nice if you can put your thoughts on that bug, since to someone who sees it now it's on the verge of WONTFIX [04:59:10] (+ i've a patch -2'd on that one's resolution) [04:59:29] OK, I will [04:59:36] thank you :) [05:56:39] New patchset: Ori.livneh; "Initial commit" [analytics/blog] (master) - https://gerrit.wikimedia.org/r/60793 [05:56:59] Change merged: Ori.livneh; [analytics/blog] (master) - https://gerrit.wikimedia.org/r/60793 [06:00:30] New patchset: Ori.livneh; "Script for reporting daily activity on Wikimedia blog" [analytics/blog] (master) - https://gerrit.wikimedia.org/r/60794 [06:05:08] New review: Ori.livneh; "@Milimetric: this script adds some data points that Tilman requires. I'm merging it to provide some ..." [analytics/blog] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/60794 [06:05:09] Change merged: Ori.livneh; [analytics/blog] (master) - https://gerrit.wikimedia.org/r/60794 [06:25:33] New review: Ori.livneh; "\o/ thanks!" [analytics/glass] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/60704 [06:25:33] Change merged: Ori.livneh; [analytics/glass] (master) - https://gerrit.wikimedia.org/r/60704 [13:50:39] moooorning [13:50:50] milimetric: i think i need some help with the user-metrics dev env [13:50:54] just looked at the card again [13:50:55] https://mingle.corp.wikimedia.org/projects/analytics/cards/570 [13:51:03] there's a bunch of stuff about creating scripts to generate fake data [14:04:26] morning guys [14:08:47] morning [14:17:07] milimetric morning [14:18:16] morning [14:18:36] morning [14:18:41] oh! [14:18:45] sorry ottomata, didn't see [14:18:49] some folks are talking about settting up limn in wikimedi-labs, not sure if you saw that [14:19:02] ok, I can switch on that since we're done with 388 [14:19:13] no, didn't see that [14:19:13] what's wikimedia-labs? [14:19:43] oh! the irc channel [14:20:14] yup [14:21:18] so ottomata, which task are you at right now? Generating fake data? I thought we said we'd write a script to import an xml dump [14:21:35] whhaaaaat? [14:22:47] milimetric: i'm basically done with the puppetization [14:22:52] waiting for ori's review [14:23:00] but i suppose you can checkout the change before it is approved and try it if you want [14:23:14] but, since you (and drdee) know mediawiki and user_metrics better than me [14:23:21] is this about the user metrics api import of a mw dump? [14:23:21] i thought you guys would know what to do better for the generation of the fake data [14:23:24] user metrics [14:23:28] because then use mwdumper [14:23:37] and download a dump file [14:23:40] see mingle card [14:23:43] that's fast [14:23:45] and requires no coding [14:24:23] mingle card says: [14:24:23] Generate script to seed MW database with random data for the following tables: [14:25:13] 1 sec [14:25:39] MW Core Tables [14:25:40] revision and page (a small wiki from dumps.wikimedia.org can be used; use mwdumper not Special:Import or importDump.php) [14:25:51] mwdumper is a java program [14:26:56] yeah, drdee, the card's bad in this instance [14:27:08] we should update with the better/easier plan to use the mwdumper [14:28:09] https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/tools/mwdumper [14:28:19] you guys always beat up my cards :( [14:28:46] it's true, you're like the card doctor [14:28:55] nobody ever comes to you to say how great their card is doing [14:29:06] it's only - doc, this card broke its ankle [14:29:07] right :( [14:29:57] ottomata: so if someone wanted to set up a Limn instance, what would be the easiest way to do that with your puppet thing? [14:30:00] it is easy :) [14:30:02] considering it's not in that config page [14:30:05] 1) download xml dump file [14:30:23] 2) clone mwdumper; create jar using maven (i fixed the pom) [14:30:44] 3) run mwdumper, specify input file, destination database, user credentials [14:30:55] 4) done, drink coffee give drdee a hug [14:31:06] * milimetric hugs drdee pre-emptively [14:31:25] this will seed the revision and page tables in mediawiki [14:31:28] milimetric: should I chime in in #labs? [14:31:40] yeah, sure [14:32:21] one thing about limn; it seems that kripke is becoming the default host for limn instances and that worries me [14:32:27] more folks are getting access to that box [14:32:42] we need to fix this real soon now [14:32:51] well, we're gonna do this particular new instance somewhere else [14:33:17] but e2 stuff is on kripke for example [14:33:31] and kaldari and benny are now admins on kripke as well [14:35:17] in related thoughts [14:35:26] we should move nexus off of labs i think…orrrrrrr at least to its own instance [14:37:40] milimetric: if that deploy thing works…let's move reportcard to stat1001! [14:37:41] yes [14:37:44] we should [14:37:58] well, the problem with that was we didn't have authentication [14:38:02] now we have authentication though [14:38:03] hm... [14:38:42] well, the other thing is Chris hasn't done the code review for Limn yet [14:38:45] we've got a meeting with him today [14:38:46] ah ok [14:38:48] https://mingle.corp.wikimedia.org/projects/analytics/cards/571 [14:38:48] ok let's leave it then [14:39:01] maybe let's wait for his first impressions and do it afterwards [14:39:18] yeah, drdee, i'd be ok with not puppetizing nexus if we just moved it to another box, if you want to keep mvp [14:39:27] we shoudl puppetize eventually [14:39:34] are you now throwing mvp at me? [14:39:38] hahah [14:39:42] ahhhhhhh the shame [14:39:44] I just did [14:39:47] shame to you [14:39:51] so embarrased [14:39:59] i'm going to go wash my mouth out with soap [14:40:07] you should [14:40:21] at least I didn't say 'synergize' like a certain someone else in this chatroom did in the review meeting yesterday [14:40:35] WHO WAS THAT? [14:40:50] are we all getting fancy consultancy speak [14:40:51] ? [14:41:00] someone who apparently doesn't like the imperial system of measurement.... [14:41:14] well, ok that is probably all 3 of us [14:41:19] but it wasn't me and it wasn't you [14:41:34] IT WAS DAN! [14:41:51] buuuuut of course he used to cruise around in those kind of environment [14:41:52] s [14:41:56] :D :D :D [14:43:02] that's true [14:44:00] lol [14:44:12] ok, I hail the word synergize from way before it was dirtied by corporate america [14:44:35] it was part of my formative years in romanian theoretical math and physics [14:45:09] so I refuse to give it up just 'cause suits stole it from me :) [14:45:22] A+B+C+D=synergize ???? [14:48:28] no man, synergy and entropy, the two forces of the universe [14:48:47] to control them is to control all things [14:49:04] sometimes it's good to let things chaotically expand, sometimes it's good to bring them in [15:40:10] milimetric: you see, no mingle card for customer adoption :D :D :D [15:43:17] oh you mean my participation in the labs discussion? [15:43:25] yeah [15:43:27] i'm making a card for it, definitely [15:43:32] editing it as we go [15:44:10] i feel like maybe we need a new type of card? [15:44:17] "customer interfacing?" [15:44:47] are you joking? [15:45:26] sadly, no [15:45:27] :) [15:45:59] i'll put it in as customer opportunity for now maybe? [15:46:15] no; sounds more like infrastructure task [15:46:27] k [16:25:39] xyzram, average: i have switched master and field_delim_param in the udp-filters repo [16:25:47] the old master can be found in branch 'backup' [16:25:50] please pull [16:25:56] and continue development against master [16:30:48] Ok, thanks. I found a lot of potential to speed things up in the old master but the new branch already has fixed those. [16:32:37] my apologies,very sorry [16:40:59] average, around? [16:47:00] ottomata: a warning reminder about nexus [16:47:12] i was totally unable to get things to work without a public ip on the box [16:47:17] that's why nexus runs on kripke [16:47:44] i tried to set up nexus.wmflabs.org on a different instance and test with a tunnel, but i could never even access it in a browser [16:47:45] *shrug* [17:02:00] poke ottomata [17:02:01] scrum [17:02:47] average: scrum [17:26:58] drdee: lost connection, but I had one thing to add [17:27:01] drdee: https://mingle.corp.wikimedia.org/projects/analytics/cards/632 [17:28:11] thx! [17:28:28] drdee: it is done [17:28:37] ? [17:28:50] on kraken? [17:28:50] it is a card that was created after the task was finished [17:28:54] no, on stat1 [17:29:09] :( [17:29:17] but kool [17:29:19] eh [17:29:24] was this done last sprint? [17:29:33] finished yesterday at 6pm [17:29:57] do we want to demo this next sprint demo? [17:30:10] it's not interesting but you can [17:30:28] dunno [17:30:31] your call [17:31:08] i don't think it merits a demo [17:31:39] k [17:54:02] drdee, i forget! what was that cool git command that let you add only parts of diffs of files to commit? [17:58:01] ottomata: http://stackoverflow.com/questions/1085162/how-can-i-commit-only-part-of-a-file-in-git [17:58:06] git add -p [17:58:14] yeah got it [17:58:15] danke [18:37:19] ottomata, milimetric: kraken is apparently affected by this bug: https://issues.apache.org/jira/browse/PIG-3195 [18:37:26] maybe you guys can also up vote the bug :) [18:37:33] well [18:37:39] it's not that "kraken is affected" [18:37:47] it's that "random pig jobs are effected" [18:37:49] don't you love my sense of drama [18:37:50] ? [18:37:55] i was thinking [18:38:04] we're running pig 0.10 [18:38:06] i was trying to rally the troops [18:38:07] we could try 0.11 [18:38:16] nahhhh [18:38:28] let's stick to the distribution and besides the bug is open [18:40:38] i need food. [18:40:42] interested, drdee? [18:40:49] not yet [18:40:51] kk [18:46:57] drdee, did you mean to put multicast relay to gadlinium on this week? [18:47:01] not oxygen precise, right? [18:47:14] yeah, 1 sec [18:48:16] fixed [18:49:10] drdee, ottomata random question re: leslie's e-mail, is there a bastion which gives access to stat1? [18:50:34] erosen: you can access stat1 through bast1001.wikimedia.org [18:50:38] k [18:50:42] YuviPanda: need help? [18:55:19] drdee: who should I talk to about getting access to bast1001 [18:58:41] you don't have access to basti1001? [18:59:10] fenari [18:59:11] bast1 [18:59:12] bast1001 [18:59:14] should all work [19:31:50] is anyone else working on any hadoop job? milimetric, ottomata, drdee? [19:31:59] no [19:32:48] nope [19:36:07] kk [19:36:17] no, not me [19:36:22] it's next on my list i think [19:36:28] ok [19:36:29] but for now cleaning up 388 [19:36:39] otto and i are going to take down the cluster briefly [19:36:56] so ping me before you start if you get to it today [19:37:25] brb lunch [20:34:36] !log stopping hadoop to make config changes: mapreduce.task.io.sort.mb 512 -> 200, mapreduce.task.io.sort.factor 100 -> 10, jvm.reuse −1 -> 1 [20:34:38] Logged the message, Master [20:44:46] !log starting hadoop [20:44:47] Logged the message, Master [20:57:58] ottomata: eep, you scrapped my changes in PS4 [20:58:19] wahhh oops [20:58:22] sorry [20:58:26] wait you made changes/! [20:58:42] ottomata: there are all sorts of small tweaks I want to make or recommend for that module but you have a lot on your plate and this is not enabled by default. Would you like me to merge it and then we'll just defer additional changes to subsequent patches? [20:59:02] what changes did you make in ps4? [20:59:24] I don't even remember; would have to diff [20:59:33] nothing critical I don't think [20:59:37] ahh i see you got rid of my absolute paths :p [20:59:41] i prefer to use them, actually [20:59:49] umapi! [20:59:54] ottomata: because sushi [20:59:59] ungnghghgh [21:00:00] you know? [21:00:23] and look at you with your befores! [21:00:34] i thought you didn't like befores [21:00:42] do custom resource types deal with requires if you don't implement them? [21:00:45] like all other resources? [21:00:48] that's what i was nervous about [21:00:51] yes [21:00:56] require is a metaparam [21:00:59] ok, so that was a bad change on my part [21:01:06] metaparam! this is new puppet territory [21:01:06] milimetric, on my way [21:01:14] ungh, umapi [21:01:21] must we? we can call it user_metrics, no? [21:01:22] i have to run for a bit but let me know if you want me to merge your patch [21:01:25] ottomata: what vexes you? [21:01:35] ottomata: FINE [21:01:35] BUT SUSHI [21:01:38] ! [21:01:41] ok. user_metrics. [21:01:58] i don't like acronyms, and prefer long names [21:01:59] in general [21:02:01] i see, it is the name [21:02:59] but, ori, yes, i'd prefer to merge as is and then fix nickpicks later [21:06:49] https://plus.google.com/hangouts/_/838f711de7c8b86a8d1a707970fb7c617e187b7f [21:07:05] ty [21:19:44] New patchset: Stefan.petrea; "Change to cron job to adapt to stat1" [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/60683 [21:24:27] Change merged: Stefan.petrea; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/60683 [21:27:53] outy outy, laters all! [21:29:13] I made a cron job [21:29:19] but uhm, it's hanging [21:29:21] ons tat1 [21:29:25] *stat1 [21:31:26] oh yes... the .bashrc was the problem [21:33:50] trying again [21:47:58] New patchset: Diederik; "Added simple script to create admin account through CLI" [analytics/E3Analysis] (master) - https://gerrit.wikimedia.org/r/60930 [21:51:26] wonder if I can simulate the way cron runs a job [21:51:33] because apparently I have some environemnt variable problems [22:05:40] milimetic, which hangout? [22:05:45] scrum? [22:05:49] erosen [22:05:59] we're in the event hangout [22:06:04] https://plus.google.com/hangouts/_/3433463179a0a361f65dc8cea9f6415c369c5e61 [22:48:13] ok, looks like it worked [22:50:27] New patchset: Stefan.petrea; "Fixed cron script for New mobile pageviews" [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/60942 [22:50:59] Change merged: Stefan.petrea; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/60942 [22:52:38] ottomata: merged [23:01:47] drdee: cron job installed, set up on stat1 [23:15:37] erosen: can you have a look at "Added simple script to create admin account through CLI" [analytics/E3Analysis] (master) - https://gerrit.wikimedia.org/r/60930 [23:15:58] sure [23:31:36] New review: Erosen; "(1 comment)" [analytics/E3Analysis] (master) - https://gerrit.wikimedia.org/r/60930 [23:37:45] nope, it is perfect! but ryan already merged it [23:38:34] erosen ^^ [23:38:42] gotcha [23:38:50] i'll deal with it next time i use it [23:38:56] ok [23:53:31] anybody know where the docs are for the GeoIP C API, e.g. GeoIP_open() [23:53:42] 1 sec [23:54:36] go to http://dev.maxmind.com/geoip/downloadable#C_API-3 and download the jar maybe?