[00:21:33] Analytics-Tech-community-metrics, DevRel-December-2015, Easy, Google-Code-In-2015: Names on scr-contributors.html should link to corresponding people.html page - https://phabricator.wikimedia.org/T118192#1886507 (Aklapper) I'm surprised that https://github.com/Bitergia/grimoire-dashboard/blob/master... [00:24:51] Analytics-Tech-community-metrics, DevRel-December-2015, Easy, Google-Code-In-2015: Names on scr-contributors.html should link to corresponding people.html page - https://phabricator.wikimedia.org/T118192#1886519 (Aklapper) a:anmol.wassan [01:33:32] Analytics-Backlog, Wikipedia-iOS-App-Product-Backlog, Zero, hardware-requests, and 3 others: Request one server to suport piwik analytics - https://phabricator.wikimedia.org/T116312#1886690 (Dzahn) Yes, it would be in wikimedia.org in DNS. @ori but how will that system know anything about traffic... [05:23:15] Analytics-Wikistats: Page view stats: monthly most popular articles not updated - https://phabricator.wikimedia.org/T48204#1886850 (ezachte) Revisiting that page I think my comment about not being terribly essential was mostly for the 2nd, 3rd and 4th tables on that page which focus on most requested non exi... [05:45:16] Analytics-Kanban, CirrusSearch, Discovery, Discovery-Cirrus-Sprint, Patch-For-Review: Setup oozie task for adding and removing CirrusSearchRequestSet partitions in hive - https://phabricator.wikimedia.org/T117575#1886875 (Deskana) [09:10:21] Analytics-Tech-community-metrics, DevRel-December-2015, Easy, Google-Code-In-2015, Patch-For-Review: Names on scr-contributors.html should link to corresponding people.html page - https://phabricator.wikimedia.org/T118192#1887056 (Aklapper) https://github.com/Bitergia/mediawiki-dashboard/pull/76 [09:15:13] Analytics-Tech-community-metrics, DevRel-December-2015: OwlBot seems to merge random user accounts in korma user data - https://phabricator.wikimedia.org/T119755#1887063 (Lcanasdiaz) We're having issues getting the metrics with the sortinghat database. Something is broken in the load process, I'm debuggin... [11:07:56] (CR) Joal: "Change has been made so that the search engine classification upgrades initial referrer classification. Sounds good to me except maybe the" (1 comment) [analytics/refinery/source] - https://gerrit.wikimedia.org/r/247601 (https://phabricator.wikimedia.org/T115919) (owner: OliverKeyes) [11:16:02] (PS1) Joal: Update oozie default error email address [analytics/refinery] - https://gerrit.wikimedia.org/r/259654 [12:11:11] (PS1) Joal: Correct pageview oozie job [analytics/refinery] - https://gerrit.wikimedia.org/r/259661 [12:11:40] (PS1) Joal: Add oozie job extracting aqs usage statistics [analytics/refinery] - https://gerrit.wikimedia.org/r/259662 (https://phabricator.wikimedia.org/T118938) [12:21:37] Analytics-Tech-community-metrics, DevRel-December-2015, Patch-For-Review: Review/update mailing list repositories in korma - https://phabricator.wikimedia.org/T116285#1887336 (Lcanasdiaz) It's updated (at last!) Thanks for the support @aklapper :) [12:31:46] hey joal [12:31:54] you're not on vacation today I see [12:32:01] huhuhu :) [12:32:18] No. t'was yesterday :) [12:32:42] cool. in that case, I need a little help deploying the geo-breakdown job [12:32:57] Erik and I agreed on formats and everything else, I tested one more time and I self-merged it [12:33:06] (I thought you were out for the rest of the week for some reason) [12:33:09] milimetric: I have seen that yes [12:33:30] milimetric: That's great: ) [12:33:34] So deploy time it is ? [12:33:48] yes, but is it possible to teach me instead of you doing it? [12:34:04] I want to save you future work if possible [12:34:17] :) [12:34:23] Completely possible [12:34:33] sweet [12:34:54] In any case, deploys will probably fall in elukey's box sonner or later ;) [12:35:14] are they documented? I looked but not too hard [12:35:28] (this one needs a refinery-source deploy too) [12:36:36] milimetric: I have no idea about doc ... I learnt from ottomata the same way you are gonna learn from me :) [12:37:12] ok, in that case I shall make a doc at Analytics/Cluster/Deploying [12:37:46] Also milimetric since you are up, before I start gathering data from beginning of november, would you mind doulble checking with me that I have not missed too big of an interesting dimension for aqs stats usage ? [12:38:09] milimetric: if you do that, I how you one more beer [12:38:09] ofc [12:38:36] The beer thing has diminishing returns, let's make it milk [12:39:02] The task for gathering pv api stats was not really priority, but I still didn't want to have it not done with data beign lost [12:39:08] :D [12:39:34] milimetric: https://gerrit.wikimedia.org/r/#/c/259662/1/hive/aqs/create_aqs_hourly_table.hql [12:40:24] * milimetric looks [12:41:22] actually milimetric, looking for docs I found: https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Refinery [12:41:42] aha! [12:42:10] joal: the stats table looks ok to me [12:42:30] side note: we are hitting each record *way* too many times for me to be comfortable :) [12:42:53] Thanks for reviewing [12:43:04] (each record in wmf.webrequest) [12:43:17] milimetric: agreed there would be really better way to do what we currently do [12:43:28] Particularly using streaming [12:43:50] aye, sorry, that was off topic [12:43:59] i'm reading the deploy steps [12:44:06] (shall I try them? or are there caveats) [12:44:19] Analytics-Tech-community-metrics, DevRel-December-2015: OwlBot seems to merge random user accounts in korma user data - https://phabricator.wikimedia.org/T119755#1887354 (Lcanasdiaz) Ok, data is updated finally. The identities were correctly splittted, have a look at the identity 028b6b8dce6241c60c313f9... [12:44:41] milimetric: let's do it together, and review :) [12:45:50] milimetric: give a minute, I'm still off topic: http://yahooeng.tumblr.com/post/135321837876/benchmarking-streaming-computation-engines-at [12:45:53] :D [12:46:15] ooh! /me reads carefully [12:46:16] milimetric: I'd actually love to test flink :) [12:47:15] 2300 nodes!? [12:47:16] hi a-team [12:47:21] mornin' mforns [12:47:25] :] [12:48:34] hey mforns :) [12:48:39] hello! [12:48:50] mforns: I'd like to spend some time with you on spark today :) [12:49:04] joal, yes please :] [12:49:06] I need to deploy with milimetric first, then I'll have time :) [12:49:13] ok, ping me please [12:49:23] sure mforns, thx :) [12:49:41] yeah you read write milimetric : 2300 node :D [12:50:07] joal, to give you a teaser, anonymizations seems to be working, but there are problems with memory [12:50:38] mforns: backfilled yesterday's IRCs talks, so I kinda followed a bit ;) [12:50:51] ok [12:57:40] the cat just jumped on a knife and sent it flying towards me. I caught it in mid air. I both feel like a ninja and like the cat is trying to kill me [12:58:03] O.o' [12:58:22] It is no news to me you're a ninja milimetric, but I din't know about your killer-cat :) [12:58:33] hehehe [12:58:37] lol [13:03:14] milimetric: couldn't find doc about5 deploying the java code [13:03:17] joal: very interesting read [13:03:19] milimetric: We should make one [13:03:27] I'd say Flink would be my choice too [13:03:39] because it doesn't lag too far behind storm (which is more mature) [13:03:43] milimetric: the ediff between storm/flink and spark is well expained [13:03:44] but it has operator combining [13:03:51] yes [13:04:19] plus milimetric, storm is on the downside, while flink is trending [13:04:39] that shouldn't matter too much, we use Camus after all :) [13:04:57] but what should matter is how easy and fun either of these are to set up [13:05:09] because our concerns aren't as much massive performance but efficiency and time [13:05:24] true milimetric [13:05:33] One reason for flink is also scala :) [13:05:42] :) I like scala [13:05:43] a lot [13:05:45] * joal likes scala [13:05:46] so, deploy? [13:05:50] sure :) [13:05:54] k, i'm on tin [13:06:00] about to do the git deploy [13:06:02] arf, first, deploy java ! [13:06:06] W [13:06:12] oh right [13:06:26] well, doesn't matter really, not like deploying this would start anything right? [13:06:30] Which means: update changelog.md in refinery-source with deployed changes [13:06:43] ok, I'll submit a patch for that [13:06:48] great :) [13:07:33] joal: so how do you guys update this? there's some automatic way? [13:07:44] or manually look through the git log [13:07:52] manualllllll [13:08:14] milimetric: While you're doing this, I'll put up a wiki page for this process [13:08:57] joal: https://wikitech.wikimedia.org/w/index.php?title=Analytics/Cluster/Refinery-source&action=edit&redlink=1 [13:09:31] k milimetric, I was going to update refiniery, but I think you're right, anither page is probably better [13:12:26] joal: do we update something to say this is now v0.0.24? I couldn't find 0.0.23 other than in some weird xml files [13:12:41] milimetric: maven does that for us :) [13:18:55] joal: oh! I see, you're waiting for me to push this commit. Maven will do that in a *separate* commit [13:18:56] :) [13:19:08] * milimetric detected deadlock [13:19:11] Yessir, sorry it was not clear :) [13:19:13] * milimetric is smarter than a Turing machine [13:19:17] huhuhu [13:19:21] * joal is not [13:19:49] * milimetric thinks Turing machines just need the concept of "awkward pause" and they'll be smarter than him [13:20:23] (PS1) Milimetric: Update changelog.md for v0.0.24 deployment [analytics/refinery/source] - https://gerrit.wikimedia.org/r/259675 [13:21:33] milimetric: I usually ask andrew to review those changes, for this one, I'll take resposibility :) [13:22:06] (CR) Joal: [C: 2] "Looks good to me, let's deploy" [analytics/refinery/source] - https://gerrit.wikimedia.org/r/259675 (owner: Milimetric) [13:25:04] (CR) Joal: [V: 2] "Looks good to me, let's deploy" [analytics/refinery/source] - https://gerrit.wikimedia.org/r/259675 (owner: Milimetric) [13:25:25] I thought jenkins would have done that on its own ... -^ [13:26:21] awkward pause ftw again! [13:26:41] joal: jenkins only does that if it's configured. And we don't have that set up on many repos [13:27:27] Analytics-Tech-community-metrics, DevRel-January-2016: Key performance indicator: Top contributors: Should have sane Ranking algorithm which takes (un)reliability of user data into account - https://phabricator.wikimedia.org/T64221#1887419 (Aklapper) [13:29:22] milimetric: I thought it was done for refinery-source [13:34:39] joal: so now we do magic with maven? [13:38:54] milimetric: just updated the page: https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Refinery-source [13:39:15] milimetric: plus, got a phone call from my brother, sorry for high latency :S [13:39:56] np at all [13:40:02] * milimetric reads [13:40:24] joal, is 8 executors 4GB each too much? [13:40:50] mforns: not really :) [13:40:56] ok [13:41:39] mforns: our cluster has 1.6TB ram, so 32G is not that much :) [13:41:56] joal, mmmm ok [13:42:02] now I get the dimension of that :] [13:45:30] woah, weird, this pushes those commits directly to gerrit [13:47:36] joal: it says I'm unauthorized, I'm assuming 'cause I don't have passwords configured for archiva? [13:49:43] correct milimetric [13:49:51] Let me gives that to you [13:55:03] milimetric: so, how's it going ? [13:55:29] good! how are you?! [13:55:34] oh, you mean with the deployment [13:55:46] right now it's stuck on [INFO] 9753/9753 KB [13:56:00] huhu :) [13:56:07] My brother is fine, thank you :) [13:56:15] op, just went past [13:56:20] great [13:57:15] seems to be doing a bunch of file copying / building / random stuff [13:57:36] actually it builds the jars and upload them to archiva [13:57:56] milimetric: I updated https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Refinery with link to the new page [14:09:10] this thing is still building... it's crazy! [14:09:21] Still building, or uploading ? [14:09:25] it looks like it's downloaded the internet, deleted it, then downloaded it again [14:09:34] arf :( [14:09:39] both uploading and building I think [14:10:04] k [14:18:00] ok, just finished [14:18:08] how do I check archiva for the jars? [14:18:15] it looked like it uploaded a lot of different things [14:18:50] it should have uploaded 1 jar per module, which means 6 jars for us [14:19:01] milimetric: go to https://archiva.wikimedia.org/#welcome [14:19:23] I usually browse to org.wikimedia.analytics.refinery [14:20:05] Yay, 0.0.24 is here for hive :) [14:20:32] haven't found it yet :) [14:20:50] I see this... https://archiva.wikimedia.org/#artifact/org.wikimedia.analyics.refinery/refinery-hive/0.0.4 [14:20:57] which says 0.0.4...? [14:20:58] Same for core --> I don't check each of them just 2 or three [14:21:22] try 0.0.24 [14:21:38] I went here: https://archiva.wikimedia.org/#artifact/org.wikimedia.analytics.refinery.hive/refinery-hive [14:22:01] It shows you the various versions it has [14:22:21] I don't understand how you got there... when I browse I end up here: https://archiva.wikimedia.org/#browse/org.wikimedia.analyics.refinery [14:22:56] Analytics-Kanban, Patch-For-Review: Create a dedicated hive table with pageview API only requests for reporting [5 pts] {melc} - https://phabricator.wikimedia.org/T118938#1887513 (JAllemandou) [14:23:28] Ah, may that: there are 2 links when you are here: https://archiva.wikimedia.org/#browse/org.wikimedia [14:23:34] analytics, analytics.refinery [14:23:40] go to the first one :) [14:23:43] Then refinery [14:23:48] * elukey is trying to follow milimetric and joal but for the moment he didn't really understand much [14:23:57] Hi elukey :) [14:24:00] o/ [14:24:11] We are double checking that jars have been correctly uploaded to archiva [14:24:28] joal: there's a misspelled one: https://archiva.wikimedia.org/#browse/org.wikimedia.analyics.refinery [14:24:35] analyics [14:24:41] True ! [14:24:48] that bad one is typo-squatting the good one :) [14:24:52] oh yes I was joking, probably I'll need to bang my head against the whole process multiple times before getting a glimpse [14:24:56] and my mild dyslexia doesn't help :) [14:25:03] :D [14:25:07] but +1 for milimetric's documentation skills [14:25:16] Yessssss ! [14:25:54] elukey: hey, factually, you're totally wrong, I'm the worst at docs, but I'll take it! [14:26:01] :P [14:26:29] ok, joal, so now I can git deploy refinery? [14:26:33] and do the sync from stat1002? [14:26:47] not yet: need to add the jars to the refinery repo :) [14:27:07] i see [14:27:13] Analytics-Kanban, Patch-For-Review: Create a dedicated hive table with pageview API only requests for reporting [5 pts] {melc} - https://phabricator.wikimedia.org/T118938#1887531 (JAllemandou) a:JAllemandou [14:27:49] milimetric: jars to be added in the artifacts folder of the refinery repo, plus links updated [14:28:04] milimetric: Don't change the one that have not been changed (we update lazyly) [14:28:30] Like in your case, I think core, hive and camus is all that changes [14:28:33] ok, but there were no details on how to do this so I'm not sure [14:28:40] camus? [14:29:00] milimetric: you said "let's document it in a wiki page", so if your last sentence is right it'll be basically my first task to fill it right? :) [14:29:07] There is a change in camus, but it's not functional, you can forget about it :) [14:30:06] oh, joal, no that's totally ok, let's update everything that was updated otherwise we'll forget [14:30:18] ok :) [14:30:40] elukey: i think we're all trying to just make this better now: https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Refinery-source [14:31:08] so, joal, you gotta walk me through this, I'm not sure where the jars are supposed to be or anything, complete noob [14:31:28] ah! artifacts? [14:31:38] milimetric: so that I replace the XXX placeholder in the doc, let's find where the jars are on your local machine [14:31:54] There should be a folder under refinery-source [14:32:07] in artifacts I see: refinery-hive.jar -> org/wikimedia/analytics/refinery/refinery-hive-0.0.23.jar [14:33:05] yes milimetric [14:33:43] You shouold copy the jars from the special folder in refinery-source into refinery/artifacts/org/wikimedia/analytics/refinery/ [14:33:48] so the process is - download the jars and put them in the correct org/blah/blah path, then update the symlinks? [14:33:52] Then change the links [14:33:55] k [14:34:00] Then commit [14:34:03] k [14:34:08] milimetric: Let's find them locally [14:34:25] In the folder where you perfomred release [14:34:38] joal: I'm seeing ../refinery-source/target/checkout/refinery-hive/target/refinery-hive-0.0.24.jar [14:34:39] There should be a folder with a special name (can't recall it) [14:34:52] It's not target [14:34:59] It's another one I think [14:35:05] target is the one the doesn't work [14:35:12] guard? [14:35:15] nope [14:35:18] let's batcave :) [14:35:23] k [14:35:36] wait joal [14:35:43] but then elukey can't follow as easily [14:35:48] True [14:35:55] Let's just find that folder [14:36:02] sure [14:36:02] The rest is easy enough on IRc [14:36:06] here's what's in my refinery-source: [14:36:14] https://www.irccloud.com/pastebin/x9DhAnHA/ [14:36:53] looks like whatever special folder there was got deleted, so I can just download them from archiva [14:37:07] yessitr [14:37:08] that's better anyway, because it includes the double checking step [14:37:15] Cool [14:37:42] (I'll update the docs for this, I have an idea what to write to make it easy for newbs like me) [14:37:58] Cool [14:37:59] Analytics-Tech-community-metrics, DevRel-December-2015: OwlBot seems to merge random user accounts in korma user data - https://phabricator.wikimedia.org/T119755#1887552 (Aklapper) @Lcanasdiaz: I see the "[[ https://github.com/Bitergia/mediawiki-identities/commit/21ca22c956aa20ba8ca2b8add99de4d72f6c4748 |... [14:38:01] Thanks :) [14:39:53] Analytics-Tech-community-metrics, DevRel-December-2015: OwlBot seems to merge random user accounts in korma user data - https://phabricator.wikimedia.org/T119755#1887553 (Aklapper) Plus looking at http://korma.wmflabs.org/browser/scr-contributors.html it seems that several staffers lost their affiliation... [14:40:07] joal, I'll leave for 20 mins to have lunch, and come back :] [14:40:25] ok mforns, sorry, took long to deploy :) [14:40:33] np, joal :] [14:43:41] joal: it's really weird but the download links have 127.0.0.1 in them?! http://127.0.0.1:8080/repository/releases/org/wikimedia/analytics/refinery/camus/refinery-camus/0.0.24/refinery-camus-0.0.24.jar [14:44:07] Yes, milimetric, I know that --> Change that host with archiva.wikimedia.org and it works [14:44:15] I don't know why :( [14:44:24] right, I figured, just making sure you knew [14:44:39] Yeah, I know (knowing is not enough) [14:44:53] heh, well this can't possibly be an issue in all archiva versions [14:45:00] that's like ... the entire purpose of this platform [14:45:16] maybe it's just a configuration problem [14:49:23] joal: and git fat will replace the large files with a placeholder when I commit? [14:49:30] or do I have to do fancy git fat stuff [14:49:48] If you have git fat setup correctly, it should do its magic :) [14:50:11] joal: that's a big if... I remember helping andrew with it a long time ago but I'm totally unsure that it's still set up [14:50:16] how do I check? [14:50:33] hm [14:51:25] look into .git/config for refinery folder [14:52:31] joal: my .git/config looks spectacularly skinny [14:52:35] (nothing fat about it) [14:52:41] :) [14:52:57] Talking about the thing, not always being :) [14:53:04] filter "fat" ? [14:53:20] filter? [14:53:29] like cat .git/config | grep fat [14:53:29] ? [14:53:33] yes [14:53:41] yeah, no fat [14:53:46] 100% lean [14:53:51] fat free [14:53:52] Arf :( [14:54:08] :) it's ok! there must be docs for this too [14:54:09] so this is good [14:54:17] The thing is, it'll get fat (actually git-fat) should be named git-unfat) [14:54:43] milimetric: there is doc about that in the readme of refienry I think [14:54:44] (I'm actively editing the Refinery-source page as we do this) [14:54:51] ok, I'll check it out [14:54:52] Awesome milimetric :) [15:01:14] joal: I'm getting: [15:01:16] ERROR:git-fat: Error reading or parsing configfile: /home/dan/projects/refinery/.gitfat [15:01:23] because it looks like I have a .gitfat file from Jly [15:01:26] *July [15:01:28] which says: [15:01:39] [rsync] [15:01:39] remote = archiva.wikimedia.org::archiva/git-fat [15:01:39] options = --copy-links --verbose [15:01:57] Ahhhhh ! My mistake milimetric, this was the one I was looking for :( [15:02:04] Then you're good to go [15:02:10] it says error :) [15:02:12] doesn't sound good [15:02:24] (it does that when I do git fat pull) [15:02:41] (but I also didn't have git fat installed, so I pip install git-fat) [15:02:47] k [15:02:58] but the error's still bad, right? [15:03:21] Right [15:03:25] I never git fat pull [15:03:35] So I can't really say :( [15:04:07] oh ok, hm... I'm very skeptical of magic [15:04:11] but I'll commit and see what happens [15:04:29] k [15:05:10] (PS1) Milimetric: Update refinery-source jars to 0.0.24 [analytics/refinery] - https://gerrit.wikimedia.org/r/259707 [15:05:30] joal: hm... after the commit those jars are still full sized [15:05:44] milimetric: https://gerrit.wikimedia.org/r/#/c/259707/1/artifacts/org/wikimedia/analytics/refinery/refinery-camus-0.0.24.jar [15:05:47] :D [15:06:10] MAGIIIIIIIIIC (http://images.google.fr/imgres?imgurl=http://i.imgur.com/YsbKHg1.gif%3Fnoredirect&imgrefurl=http://imgur.com/gallery/YsbKHg1&h=252&w=275&tbnid=HryI-43v3FHG-M:&tbnh=147&tbnw=160&docid=oTJwQHdeAQymIM&usg=__ZpNAT6srAH8u5dIUWG0QI8o_lIs=&sa=X&ved=0ahUKEwif-svilePJAhWHnRoKHcIoCzwQ9QEIITAA [15:06:39] I agree, that should definitely pop up whenever git fat does something [15:06:50] but... the file in the repo is not changed, that's what really freaks me out [15:06:57] milimetric: I double check the jar id [15:07:11] jar doesn't change in the repo [15:07:18] it's not supposed to [15:07:32] It's just not uploaded in the git repo [15:07:38] instead, you have a git-gat link [15:09:08] k, that's fine [15:09:12] I updated https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Refinery-source [15:09:27] (CR) Ottomata: [C: 2 V: 2] Update oozie default error email address [analytics/refinery] - https://gerrit.wikimedia.org/r/259654 (owner: Joal) [15:09:27] * joal reads [15:09:28] so now when you merge that in gerrit, I can git deploy [15:09:41] yessir [15:10:38] ottomata: Hey, do you know I can check for jar ID in archiva? [15:11:25] to see if the git-fat sha you have matches the one on archiva? [15:11:31] yessir [15:11:35] hmmm [15:12:05] no, not without login to the archiva server [15:12:09] i can see it for you though :/ [15:12:18] hmmm [15:12:18] oh [15:12:18] i mean [15:12:22] np ottomata, we are gonna trust git-fat :) [15:12:23] you can calc the sha [15:12:31] if you dl the jar [15:12:32] milimetric: --^ [15:12:35] yes [15:12:47] shasum whatever.jar [15:13:00] k [15:13:27] milimetric: forgot to double check a thing before merging :( [15:14:18] joal: you haven't merged it yet... [15:14:31] milimetric, right :) [15:14:43] Was wondering about the jar version in your code [15:15:07] I'll merge the new jars in artifacts repo now [15:15:24] (CR) Joal: [C: 2 V: 2] "Moving toward deploy" [analytics/refinery] - https://gerrit.wikimedia.org/r/259707 (owner: Milimetric) [15:16:44] back joal [15:16:50] joal: but you said you forgot to double check something? [15:17:02] Doing it now :) [15:17:44] what's that? I can add it to the docs [15:18:48] milimetric: it's in the code you merged: one property missing in your new job [15:19:00] milimetric: Or at least I think [15:19:34] oh I see [15:19:59] milimetric: parameterization of jar version ;) [15:21:32] oh, joal I just copied that, you mean this: refinery-hive-${refinery_jar_version}.jar; [15:21:56] Yes milimetric q [15:22:15] It's not set in coordinator.properties IIRc [15:22:25] oh! I thought it was magically provided [15:22:28] :D [15:22:42] it's not set, yes, that needs to be hard-coded then? [15:24:01] Yeah, look at refinery/oozie/webrequest/refine/bundle.properties [15:25:00] k, patching [15:28:43] ottomata: I suggest not restarting the oozie jobs to change the email thing. I think it's better to have them restart for a meaninful change, changing that thing meanwhile [15:28:50] yeah [15:28:52] agree [15:28:54] we'll just wait [15:28:56] cool [15:29:06] thx [15:34:05] milimetric: I go try to help mforns if ok for you [15:34:13] mforns: batcave ? [15:34:31] joal, yes! [15:35:00] ok with me, I'm still working on the patch [15:36:20] (PS1) Milimetric: Fix problem referencing jar from geo job [analytics/refinery] - https://gerrit.wikimedia.org/r/259723 [16:09:06] * elukey needs to study the Refinery [16:09:46] milimetric, joal: documenting how to deploy to cluster would be awesome, please do. I would like to be able to do it myself too [16:10:02] nuria: we did with milimetric [16:10:33] joal: thank you [16:10:38] np :) [16:19:01] Analytics-Tech-community-metrics, DevRel-December-2015: OwlBot seems to merge random user accounts in korma user data - https://phabricator.wikimedia.org/T119755#1887697 (Lcanasdiaz) >>! In T119755#1887553, @Aklapper wrote: > Plus looking at http://korma.wmflabs.org/browser/scr-contributors.html it seems... [16:21:43] Analytics-Wikistats: Page view stats: monthly most popular articles not updated - https://phabricator.wikimedia.org/T48204#1887699 (Nuria) >I don't think the new page view api can zoom in on those missing pages/files in particular. True, it cannot. > But again, there seems to be little demand for it. Ok, le... [16:32:44] (CR) Nuria: "Please see @joal comment i also just have one minor comment too." (1 comment) [analytics/refinery/source] - https://gerrit.wikimedia.org/r/247601 (https://phabricator.wikimedia.org/T115919) (owner: OliverKeyes) [16:35:56] (CR) Nuria: [C: 2] Correct pageview oozie job (1 comment) [analytics/refinery] - https://gerrit.wikimedia.org/r/259661 (owner: Joal) [16:39:26] (CR) Nuria: Fix problem referencing jar from geo job (2 comments) [analytics/refinery] - https://gerrit.wikimedia.org/r/259723 (owner: Milimetric) [16:41:19] (CR) Milimetric: Fix problem referencing jar from geo job (1 comment) [analytics/refinery] - https://gerrit.wikimedia.org/r/259723 (owner: Milimetric) [16:42:47] milimetric: let's talk about jar version of oozie job on standup [16:43:29] sure [16:44:20] elukey: I have invited you to our staff meeting, if you can come gret. If not, no big deal. [16:48:12] nuria: sure! I have some free time, I'll be glad to join :) [16:48:18] excellent [17:00:16] ottomata: standuppp [17:00:43] OO [17:00:45] finding headphones [17:14:56] joal: can you post the flink link here.... I'm not in the hangout [17:15:01] hmsure [17:15:07] http://yahooeng.tumblr.com/post/135321837876/benchmarking-streaming-computation-engines-atb [17:24:43] nuria: when you +2, code is not merged - Did you meant to merge or not ? [17:25:04] joal: i was waiting for ottomata to look at it but it can be merged yes [17:25:18] k nuria [17:25:23] which? [17:26:09] (CR) Nuria: [C: 2 V: 2] Fix problem referencing jar from geo job [analytics/refinery] - https://gerrit.wikimedia.org/r/259723 (owner: Milimetric) [17:26:39] ottomata: https://gerrit.wikimedia.org/r/#/c/259661/ [17:32:35] nuria: not sure if you guys are already in the batcave but I think that I need permissions to join [17:32:36] elukey: can you join: https://plus.google.com/hangouts/_/wikimedia.org/a-batcave [17:33:10] elukey: i have sent you a new invite [17:33:44]