[00:50:46] 10Quarry, 10Cloud-Services, 10Community-Wikimetrics, 10DBA, and 2 others: Evaluate future of wmf puppet module "mysql" - https://phabricator.wikimedia.org/T165625#4071489 (10Dzahn) stopped using on in mediawiki_deployment_server role: https://gerrit.wikimedia.org/r/#/c/421197/ [02:40:06] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Clean up XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071575 (10Krinkle) [02:40:43] 10Analytics, 10MediaWiki-API, 10Wikipedia-Android-App-Backlog, 10Wikipedia-iOS-App-Backlog: Add page_id and namespace to X-Analytics header in App / api requests - https://phabricator.wikimedia.org/T92875#1122344 (10Krinkle) 05Open>03Resolved a:03Krinkle This is working and documented at 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071593 (10Krinkle) [02:53:20] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071594 (10Krinkle) [03:07:23] 10Analytics, 10MediaWiki-API, 10Wikipedia-Android-App-Backlog, 10Wikipedia-iOS-App-Backlog: Add page_id and namespace to X-Analytics header in App / api requests - https://phabricator.wikimedia.org/T92875#4071616 (10Ottomata) Wait, really, so we get `page_id` and `ns` set in API requests now? I thought th... [03:09:28] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071633 (10Krinkle) [04:23:38] 10Analytics, 10MediaWiki-API, 10Wikipedia-Android-App-Backlog, 10Wikipedia-iOS-App-Backlog: Add page_id and namespace to X-Analytics header in App / api requests - https://phabricator.wikimedia.org/T92875#4071672 (10Krinkle) @Ottomata The vast majority of API modules either don't relate to a page, or (as y... [04:24:36] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071673 (10Krinkle) [04:25:52] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071575 (10Krinkle) [05:01:00] 10Analytics, 10MobileFrontend, 10Performance-Team, 10Readers-Web-Backlog, 10Technical-Debt: Figure out XAnalytics stuff - https://phabricator.wikimedia.org/T190381#4071701 (10Krinkle) [05:27:42] (03PS11) 10Nuria: Create and manipulate date objects according to UTC timezone [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/417476 (https://phabricator.wikimedia.org/T189266) (owner: 10Fdans) [05:31:45] (03PS12) 10Nuria: Create and manipulate date objects according to UTC timezone [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/417476 (https://phabricator.wikimedia.org/T189266) (owner: 10Fdans) [05:37:17] (03CR) 10Nuria: ">WIth this change, for me right now all the dashboard metrics still show January as last month, >while they should be February right?" [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/417476 (https://phabricator.wikimedia.org/T189266) (owner: 10Fdans) [05:38:06] (03CR) 10Nuria: Create and manipulate date objects according to UTC timezone (031 comment) [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/417476 (https://phabricator.wikimedia.org/T189266) (owner: 10Fdans) [07:31:37] 10Analytics, 10Analytics-EventLogging, 10Patch-For-Review, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#4071959 (10elukey) I was able to manually run the eventlogging_cleaner but with some tweaks (like https://gerrit.wiki... [08:08:44] Hi elukey [08:08:55] morningggg [08:09:57] elukey: if you want I can show you the notebook I was playing with yesterday, and tell you how to replicate [08:10:53] joal: mind if we do it after lunch? (if you have time) [08:11:25] elukey: I'll have time, but will flexibility as I'll need to bring Naé to the doctor [08:12:03] of course :) [08:29:37] 10Analytics-Kanban, 10Patch-For-Review, 10User-Elukey: Add the prometheus jmx agent to AQS Cassandra - https://phabricator.wikimedia.org/T184795#4072047 (10elukey) [08:42:53] The man is crazy ... https://xkcd.com/1970/large/ [08:46:22] oh my.. what's that?? [09:07:15] joal: http://mail.openjdk.java.net/pipermail/announce/2018-March/000247.html [09:07:54] let's upgrade!! [09:07:56] :D [09:09:04] Yay elukey! I love the idea of consolidating the JDK forest :D [09:12:39] the Java situation will become worse forest-wise [09:12:57] they're moving to more frequent releases (which less changes in between) [09:13:16] but only a few of those are long term supported and suitable for distros [09:13:36] 10 will see a fairly short support time frame, so only 11 will end up distros [09:14:13] 10 is ofc in Debian right now: https://packages.qa.debian.org/o/openjdk-10.html [09:14:20] but that's interim [09:19:09] nice to know! [09:19:33] in the meantime, lemme roll restart some openjdk-8 daemons :) [09:26:44] moritzm: afaics the new openjdk-8 is not deployed on all the analytics* hosts right ? [09:28:20] whenever you have time I can roll restart hadoop and test one of the kafka hosts (like kafka-jumbo1001) [09:30:33] I forgot, did we upgrade one Hadoop last week? [09:30:53] or I did only want to suggest that to you? :-) [09:31:03] (Hadoop node) [09:35:04] moritzm: we did 1070 iirc, but then nothing more (like a kafka one etc..) [09:35:29] I can upgrade those packages myself, didn't want to step into your feets while upgrading :) [09:39:52] I can do that, I have the debdeploy spec ready any, upgrading the Hadoop cluster now [09:41:43] yeah, confirmed, 1070 runs the new Java since March 19 [09:45:41] super, if you want I can start using your debdeploy procedure and free you from this task (it doesn't change much for me, just don't want to annoy you or interrupt your work to ask for upgrades :) [09:50:56] takes just a minute, I just upgraded the hadoop cluster [09:51:10] all right then! [09:51:37] jmxtrans seems like a candidate for our new wmf-auto-restart framework, it's entirely stateless, right? [09:52:26] yep it is afaik, buuut I have almost removed it from all the analytics infrastructure (zookeeper is the only one left using it IIRC) [09:52:55] (and probably the kafka hosts still not upgraded to 1.0) [09:56:14] started the rolling restart of all the node managers [09:56:27] (batch of two hosts, sleep 120s) [09:56:46] it will take a bit buut we are not in a hurry, and cumin does the work for me :) [09:57:14] the hdfs restarts will likely be a bit more invasive [09:57:25] ah, I see the problem: [09:57:50] jmxtrans is incompletely removed, it's in "rc" status, so all conffiles are retained [09:58:07] and as such the sysvinit script [09:58:17] which gets picked up by systemd [09:58:44] that's why it showed up in "systemctl list-units" for me on the Hadoop node I checked [09:59:35] running "dpkg --purge jmxtrans" via cumin across the hadoop cluster will fix that [09:59:41] ah right I should've probably run apt-get purge ? [10:00:17] when removed a single package without reverse deps I tend to use dpkg rather [10:01:16] apt-get purge is fine if all the packages on a system were installed via apt [10:01:50] but since we also use puppet people tend to deploy scripts e.g. which use a package without declaring a dependency on the package manager leve [10:02:04] and that causes problems if the underlying package is removed [10:02:10] we had that when we dropped salt [10:02:39] since it installed a dozen of Python packages people then took for granted since the Salt minion was universally available [10:03:09] but it's not a problem for jmxtrans which probably doesn't have any dependencies, so apt purge would have worked as well [10:03:58] thanks for the explanation :) [10:04:14] just run dpkg --purge jmxtrans on an1028, all fine, I'll apply it to all the cluster [10:05:58] thx [10:10:16] done! [10:27:03] hi team! [10:27:17] hi mforns [10:29:21] elukey: I upgraded kafka-jumbo1001 for the new JDK, we can use that as a canary before proceeding with the update for the others [10:30:58] ack! [10:33:18] interesting, the update of openjdk makes some changes to configs which we seem to modify, let me make a task [10:35:16] moritzm: ? [10:35:31] should I be worried for the rolling restart? [10:35:45] no, seems to only hapen for kafka-jumbo1001 [10:35:52] didn't see that kind of change for hadoop [10:37:17] on kafka-jumbo1001 /etc/java-8-openjdk/security/java.security is managed by puppet [10:37:40] and that's not the case of the hadoop cluster [10:37:54] the way Debian config handling works is: [10:38:48] if a file marked as a config file hasn't been modified and the default updated in the package changed, the config file is updated to the new package default [10:38:55] as it happens on hadoop [10:39:33] but on kafka-jumbo we're shipping a different security policy, so we need to review which changes which happened in u162 need to be incorporated into our puppetised template [10:39:58] since changes done by upstream in a stable release of Java 8 are usually important [10:42:19] ah yes right! For the TLS security review! [10:42:33] this is a good point to keep in mind for the future [10:46:42] moritzm: https://phabricator.wikimedia.org/T182993 [10:48:14] created https://phabricator.wikimedia.org/T190400 [10:48:26] not sure which analytics projects to add, please amend :-) [10:49:58] when upgrading kafka-jumbo1001 I retained the current config, so we can generally restart it too test whether it works fine with Java u162, but let's hold the complete cluster restart until T190400 is resolved [10:49:58] T190400: Review changes to /etc/java-8-openjdk/security/java.security in Kafka from u162 update - https://phabricator.wikimedia.org/T190400 [10:50:37] I can have a look at 190400 tomorrow as a second pair of eyes, but also needs to be reviewed by someone more familiar with kafka [10:53:33] moritzm: I'll have a chat with Andrew todaay [10:58:16] ok [11:22:30] a-team: FYI I am rolling restart all the hadoop hdfs datanodes for jvm updates, 1 every two minutes [11:22:40] so if you see job failures it might be me [11:22:46] roger [11:26:20] the super nice thing about this is that we are going to get also the tolerance to failed volumes (2 as opposed to 0) [11:26:38] so in case of a disk failure the node will simply replicate the data elsewhere and keep going [11:45:06] mforns: o/ [11:45:10] hey elukey [11:45:13] whenever you have time can we chat about https://gerrit.wikimedia.org/r/#/c/421233/ ? [11:45:42] elukey, sure, now, do you want to batcave? [11:45:47] or just me to review? [11:46:18] in here is fine, just wanted to chat about this (hopefully last) corner case [11:47:21] cleaner code looks good! [11:47:52] puppet code... looks awesome to me in my ignorance... [11:47:55] :] [11:48:30] the important part is that you are ok to skip those sanity checks in labs [11:48:59] sure [11:49:19] ack! This morning I've run a sanitization in beta, all good! [11:49:25] so after this we should be ready [11:59:43] elukey, I will test there today [11:59:56] the new changes with the YAML white-list [12:01:35] mforns: how are you going to do it ? Manually or do you want me to apply your patch to the deployment-prep's puppet master? [12:01:59] elukey, I haven't thought of it yet... [12:02:57] elukey, I guess... I can do that manually, it's labs so I have permits to overwrite el cleaner there, right? [12:03:01] mforns: maybe at first we could just disable puppet and change manually, but then I'd prefer to do a test to make sure that when we apply puppet the change does what it expected to do [12:03:09] you should yes [12:03:19] but disable puppet first, otherwise it'll be rewritten [12:03:39] ok, sure, when I'm finished I'll ping you to apply the patch to deployment-prep [12:03:43] ok [12:17:08] 10Analytics-Kanban, 10Patch-For-Review: Remove sensitive fields from whitelist for QuickSurvey schemas (end of Q2) - https://phabricator.wikimedia.org/T174386#4072461 (10mforns) @elukey Queries look good to me! Yea, we can remove the `where timestamp < ...` clauses. I hope the updates do not take days... +2 [12:23:50] 10Analytics-Kanban, 10Google-Summer-of-Code (2018): [Analytics] Improvements to Wikistats2 front-end - https://phabricator.wikimedia.org/T189210#4072475 (10mforns) Hi @Jibin2706, the source code is in the task's description (reference [4]). [12:25:54] joal, yt? [12:26:17] would you be ok with https://phabricator.wikimedia.org/T187427 -> 'Unknown' in the end? [12:30:15] mforns: deployment-eventlog05 runs now with the --no-whitelist-sanitization, works fine! [12:30:21] (I am running it manually) [12:30:35] elukey, awesome :] [12:31:05] 10Analytics, 10Analytics-EventLogging, 10Patch-For-Review, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#4072483 (10elukey) [12:31:25] 10Analytics, 10Analytics-EventLogging, 10Analytics-Kanban, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#3457322 (10elukey) [12:31:51] 10Analytics, 10Analytics-EventLogging, 10Analytics-Kanban, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#3457322 (10elukey) Added the `--no-whitelist-sanity-check` to deployment-eventlog05, now everything works as expected. [12:43:59] (03CR) 10Mforns: [V: 032 C: 032] "Great :]" [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/419977 (https://phabricator.wikimedia.org/T188208) (owner: 10Sahil505) [12:45:05] 10Analytics, 10Analytics-Wikistats, 10Patch-For-Review: The alert message about adblocker is not fully shown on smaller screens - https://phabricator.wikimedia.org/T188208#4072533 (10mforns) a:03sahil505 [12:46:03] 10Analytics-Kanban, 10Analytics-Wikistats, 10Patch-For-Review: The alert message about adblocker is not fully shown on smaller screens - https://phabricator.wikimedia.org/T188208#3999726 (10mforns) [12:47:06] 10Analytics-Tech-community-metrics, 10Developer-Relations (Jan-Mar-2018): Investigate how to identify our code contributors to on-wiki code (gadgets, templates, modules) on a WMF site - https://phabricator.wikimedia.org/T190164#4072545 (10Aklapper) [12:47:30] 10Analytics-Tech-community-metrics, 10Developer-Relations (Jan-Mar-2018): Investigate how to identify our code contributors to on-wiki code (gadgets, templates, modules) on a WMF site - https://phabricator.wikimedia.org/T190164#4065066 (10Aklapper) 05Open>03Resolved a:03Aklapper Works for me: {P6883} [13:04:09] * elukey afk for a bit! [13:21:03] elukey: whenever you have a couple mins we can run the queries for quicksurvey, since mforns has okayed them [13:21:04] joal: ready to merge https://gerrit.wikimedia.org/r/#/c/413265/? [13:35:50] fdans: sure [13:38:49] 10Analytics, 10MediaWiki-API, 10Wikipedia-Android-App-Backlog, 10Wikipedia-iOS-App-Backlog: Add page_id and namespace to X-Analytics header in App / api requests - https://phabricator.wikimedia.org/T92875#4072675 (10Ottomata) Great, thanks! It might nice if in the future we were somehow able to associate... [13:39:08] joal: , yt? [13:39:11] can we talk notebooks? [13:40:09] OH SHOOT joal, in goals meeting we didn't talk spark 2 [13:40:10] AH! [13:40:21] elukey: cave? [13:41:00] fdans: not now if you don't mind, a bit busy [13:41:06] sure, sorry [13:41:10] how about in 1h?? [13:41:19] yep that works! [13:41:29] ah no snap I have tons of meetings today [13:41:30] * elukey cries [13:41:40] fdans: 20mins! [13:42:39] coool! elukey [13:59:45] 10Analytics, 10Analytics-EventLogging, 10Analytics-Kanban, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#4072738 (10elukey) Updated https://wikitech.wikimedia.org/wiki/Analytics/Systems/EventLogging/TestingOnBetaCluster#Da... [13:59:55] Hey lads - Here I am :) [13:59:57] fdans: ok I am free if you are [14:00:36] milimetric: last review, and then yes, ready ! [14:00:45] k [14:00:55] ottomata: spark2 indeed - forgot yesterday [14:01:00] ottomata: tonight ? [14:01:25] milimetric: I'd be interested to know what was wrong with git on stat1005 [14:01:43] oh yeah, lemme check [14:02:40] joal: core.autocrlf=true I think [14:02:45] it's set globally [14:02:46] joal: toinght for notebook? [14:02:52] at yerr leisure :) [14:03:06] ottomata: tonight during groskin for spark2 - now for notebooks if you wish :) [14:03:17] oho ok [14:03:17] ya [14:03:45] ok joal, so you say you found a new python lib dir in your venv [14:03:47] what does that mean? [14:04:09] joal: it's possible it's set up like that for wikistats 1.0 because erik works on windows? [14:04:09] ottomata: on notebook, in a terminal, I was looking into the venv folder [14:04:23] milimetric: hm - That might !@ [14:04:33] I remember he had a lot of trouble with line endings and the repo would always show everything as changed all the time [14:04:41] yes [14:04:49] but what does that mean for use? [14:04:58] do we need to manually reinstall all deps into each venv? [14:05:04] from the python 3.4 dir? [14:05:14] yessir [14:05:16] ottomata: -^ [14:05:19] oof ok [14:05:35] ottomata: Than means any lib installed under p3.4 is unavailable for p3.5 [14:05:41] did not realize that [14:05:42] ok [14:05:47] lemme see if i can figure that out [14:05:54] ottomata: I found a solution for sasl anf hive, but it;s super hacky [14:06:03] what's that? [14:06:38] couldn't pip install sasl,but installed puresasl instead (installation succeeded) [14:07:06] I hada to downgrade thrift-sasl to 0.2.1 instead of 0.3.0, and manually patch 2 lines [14:07:38] milimetric: just double checked the druid datasource for geowiki - 1 segment of 52Mb - that's great :) [14:08:50] milimetric: Shall I merge then ? [14:09:29] joal: yeah, thanks! [14:09:51] I'll make a task for the rest of the work around this [14:10:06] milimetric: sounds good I was about to ask about the load part of it :) [14:10:16] (03CR) 10Joal: [V: 032 C: 032] "LGTM ! Merging" [analytics/refinery] - 10https://gerrit.wikimedia.org/r/413265 (https://phabricator.wikimedia.org/T188113) (owner: 10Milimetric) [14:13:23] ottomata: except from that hack and the need to reinstall some packages, the things works great for me :) [14:15:38] 10Analytics-Kanban: Checklist for geowiki pipeline - https://phabricator.wikimedia.org/T190409#4072764 (10Milimetric) [14:16:07] 10Analytics-Kanban: Checklist for geowiki pipeline - https://phabricator.wikimedia.org/T190409#4072776 (10Milimetric) [14:16:20] joal: ^ that's the task, see if I missed anything [14:17:19] joal: just curious, why not pyhive? [14:17:25] restarting the hdfs journal nodes for jvm updates [14:17:29] (btw, I am fixing, we need some dev libs installed) [14:21:43] ottomata: hmmm - I don't know why not pyhive! [14:22:03] ottomata: maybe it was not in as good shape as impyla when I started using notebooks? [14:22:20] aye [14:22:26] ok, i think i'm going to make it installed by default :) [14:22:30] elukey: I'm sorryyyyy I had to run out of my place for a bit, any chance you still free? [14:22:47] ottomata: pyhive or impyla? [14:23:06] pyhive [14:23:13] ok [14:23:19] I'm gonna need some tutos :) [14:23:25] Or make it [14:23:41] fdans: yep! [14:23:52] hmm, actually, i can do both? [14:23:56] omw elukey [14:25:28] ottomata: looking at github, pyhive seems more recent than impyla, and currently having a lo more acticity [14:25:36] ay [14:25:37] e [14:26:33] hey joal, in the end, would you be ok with https://phabricator.wikimedia.org/T187427 -> 'Unknown'? [14:26:48] milimetric: I think there is one step not done in your list which is to real-delete old data on druid (removing segments from historical doesn't delete them from deep-storage) [14:26:59] milimetric: or maybe we wanna keep them in there? [14:27:19] joal: no, we wanna delete them, but I thought we had something in place for that kind of thing? [14:27:31] milimetric: I think too - need to check [14:27:46] mforns: Hi :) Ok for me - commenting on task :) [14:27:57] thans joal :] [14:28:59] 10Analytics, 10Analytics-Wikistats: Change '--' to something more helpful in Wikistats page views by country table view - https://phabricator.wikimedia.org/T187427#4072821 (10JAllemandou) Thanks for making it clearer in my mind :) Indeed, for country names and in fronend, please feel free to use `Unknown`. Sor... [14:30:05] milimetric: we have script: refinery/bin/drop-druid-deep-storage-data - Let's just make sure we don't forget to add a cron for it :) [14:30:25] editing the task, thanks joal [14:30:50] 10Analytics-Kanban: Checklist for geowiki pipeline - https://phabricator.wikimedia.org/T190409#4072835 (10Milimetric) [14:30:54] Hi dsaez - Are you around? [14:36:32] (03PS1) 10Ottomata: Update wheels with pyhive and impyla for default Hive access in prod [analytics/jupyterhub/deploy] - 10https://gerrit.wikimedia.org/r/421298 (https://phabricator.wikimedia.org/T183145) [14:36:54] joal: were your old notebooks from notebook1001 in place/viewable/useable on notebook1003? [14:37:12] (03CR) 10Ottomata: [V: 032 C: 032] Update wheels with pyhive and impyla for default Hive access in prod [analytics/jupyterhub/deploy] - 10https://gerrit.wikimedia.org/r/421298 (https://phabricator.wikimedia.org/T183145) (owner: 10Ottomata) [14:37:22] ottomata: except from the dependencies, very much ottomata - looked the same old to me :) [14:37:49] ok great [14:37:54] so you could read them / see them [14:37:56] but maybe not run them [14:38:50] joal: did you try jupyterlab? :D [14:38:51] ottomata: If you want to be nice to users, let's also install findspark please :) [14:38:56] ottomata: I didn't ! [14:38:58] ottomata: how|/. [14:39:00] ? [14:39:05] findspark?! [14:39:19] ottomata: and pandas, numpy, scipy, matplotlib etc [14:39:24] sure [14:39:34] joal gimme big list [14:39:44] ottomata: sure [14:40:02] some im ight install from debs if they are avail [14:40:27] oh wait, we get a lot of that [14:40:31] from statistics::packages already, no? [14:40:46] ya we do [14:40:50] oh but not python3 [14:40:51] on it. [14:40:51] :) [14:41:00] ottomata: very possible - didn't check - But they were not on notebooks for sure :) [14:41:09] because not python3! [14:42:35] 10Analytics, 10Analytics-Wikistats: Change '--' to something more helpful in Wikistats page views by country table view - https://phabricator.wikimedia.org/T187427#4072870 (10mforns) Cool! @sahil505, feel free to add a change to Wikistats in Gerrit for this issue, as you did with the other task. This should be... [14:45:31] ottomata: pydruid can be added to the list - We had issues with it however - will test first [14:49:12] ottomata: same issue on pydruid seems - related to proxies (can't recall precisely - I made a PR about that a while ago) [14:50:27] ok, let's let that one be for now then [14:50:31] low hanging fruit only atm :) [14:50:44] ottomata: +1 [14:58:16] 10Analytics-Kanban, 10Google-Summer-of-Code (2018): Proposal : [Analytics] Improvements to Wikistats2 front-end - https://phabricator.wikimedia.org/T189964#4072892 (10mforns) Hi @sahil505! Thanks a lot for your proposal! Here's some feedback, I hope it helps. **Profile, synopsis and benefits** Looks great :-]... [15:04:24] (03CR) 10Fdans: [V: 032 C: 032] "Tested successfully in mac Chrome and Firefox, merging!" [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/417476 (https://phabricator.wikimedia.org/T189266) (owner: 10Fdans) [15:06:56] elukey: shall we merge this? [15:06:57] https://gerrit.wikimedia.org/r/#/c/420346/ [15:07:18] dsaez: ping again? [15:08:16] (03CR) 10Elukey: Add hi.wikimedia, zh.wikidata and wikipedia.commons to whitelist (031 comment) [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 (owner: 10Fdans) [15:09:05] (03CR) 10Fdans: Add hi.wikimedia, zh.wikidata and wikipedia.commons to whitelist (031 comment) [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 (owner: 10Fdans) [15:10:37] (03PS4) 10Fdans: Add hi.wikimedia, zh.wikidata and wikipedia.commons to whitelist [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 [15:10:39] (03PS1) 10Fdans: Release 2.1.12 [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/421311 [15:11:05] (03CR) 10Fdans: [V: 032 C: 032] Release 2.1.12 [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/421311 (owner: 10Fdans) [15:11:08] (03CR) 10Elukey: [C: 031] Add hi.wikimedia, zh.wikidata and wikipedia.commons to whitelist [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 (owner: 10Fdans) [15:13:24] (03PS1) 10Fdans: Release 2.1.12 [analytics/wikistats2] (release) - 10https://gerrit.wikimedia.org/r/421312 [15:13:38] (03CR) 10Fdans: [V: 032 C: 032] Release 2.1.12 [analytics/wikistats2] (release) - 10https://gerrit.wikimedia.org/r/421312 (owner: 10Fdans) [15:17:14] (03PS1) 10Fdans: Release 2.1.12 (for reals this time) [analytics/wikistats2] (release) - 10https://gerrit.wikimedia.org/r/421313 [15:17:38] hm - I have wonder here about the wikipedia.commons whitelist [15:17:49] fdans: http://wikipedia.commons.org [15:18:37] (03CR) 10Fdans: [V: 032 C: 032] Release 2.1.12 (for reals this time) [analytics/wikistats2] (release) - 10https://gerrit.wikimedia.org/r/421313 (owner: 10Fdans) [15:20:42] A-team - I'm gonna miss standup again today - Next week time will have changed in Europe as well, and I'll be there every day [15:21:02] I'll however be here for groskin (I'm assuming we're gonna do apss over the goals again?) [15:21:52] joal: commons.wikipedia was reported by the following query: [15:21:54] https://www.irccloud.com/pastebin/zuFMdpog/ [15:22:47] fdans: commons.wikipedia.org --> Fine ! wikipedia.commons.org -- Not fine :) [15:23:07] fdans: I should have checked the code instead of just the commi message :) [15:23:09] omg what is happening to me today [15:23:34] oh but the code is alright joal? [15:23:45] fdans: It took me three times looking at it to actually understand why I had looked at it again [15:23:56] Didn't check fdans [15:23:57] thank god, I thought I had gone on FULL DUMB mode today [15:24:01] I'm assuming so? [15:24:31] fdans: Code is correct :) [15:24:32] yeah it's all good [15:25:23] And actually commons.wikipedia is a redirect to commons.wikimedia [15:25:49] (03CR) 10Joal: [C: 031] "One nit in commit message, +1 since minor" (031 comment) [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 (owner: 10Fdans) [15:31:45] (03PS5) 10Fdans: Add hi.wikimedia, zh.wikidata and commons.wikipedia to whitelist [analytics/refinery] - 10https://gerrit.wikimedia.org/r/420346 [15:47:49] 10Analytics-Kanban, 10Analytics-Wikistats, 10Patch-For-Review: The alert message about adblocker is not fully shown on smaller screens - https://phabricator.wikimedia.org/T188208#4073063 (10sahil505) @mforns Thanks for the review :) [15:49:20] 10Analytics, 10Analytics-Wikistats: Change '--' to something more helpful in Wikistats page views by country table view - https://phabricator.wikimedia.org/T187427#4073070 (10sahil505) @JAllemandou Thanks a lot. @mforns On it :-) [16:01:58] ping joal ottomata [16:02:06] ping fdans [16:03:11] holaa ottomata ! [16:03:16] (03PS1) 10Sahil505: Changed '--' to 'Unknown' in Wikistats page views by country table view [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/421330 (https://phabricator.wikimedia.org/T187427) [16:04:23] AHHH [16:06:24] 10Analytics-Kanban, 10Patch-For-Review: Limit length of table and add "Load more rows" button - https://phabricator.wikimedia.org/T188953#4025130 (10Nuria) [16:07:17] 10Analytics, 10Analytics-Wikistats: Routing code allows invalid routes - https://phabricator.wikimedia.org/T188792#4073142 (10Nuria) [16:07:52] 10Analytics: Replace all hover actions with touch-compatible UI - https://phabricator.wikimedia.org/T188277#4073145 (10Nuria) [16:09:27] 10Analytics, 10Analytics-Cluster: Automate refinery jar cleanup - https://phabricator.wikimedia.org/T159337#4073148 (10Nuria) [16:10:16] 10Analytics, 10Analytics-Wikistats: Wikistats 2.0: Page heading style varies - https://phabricator.wikimedia.org/T187412#4073152 (10Nuria) [16:11:09] 10Analytics, 10Analytics-Kanban: Update html language for per-domain uniques - https://phabricator.wikimedia.org/T168477#4073158 (10Nuria) [16:11:18] 10Analytics: Update html language for per-domain uniques - https://phabricator.wikimedia.org/T168477#3365554 (10Nuria) [16:12:38] 10Analytics: Investigate AQS cassandra schema hash warninga - https://phabricator.wikimedia.org/T178832#4073163 (10Nuria) [16:14:17] 10Analytics: Some metrics don't work in the topic selector - https://phabricator.wikimedia.org/T188268#4073171 (10Nuria) [16:17:39] 10Analytics: We should prevent the user from trying to rediscover America - https://phabricator.wikimedia.org/T187452#4073186 (10Nuria) [16:18:06] 10Analytics, 10Analytics-EventLogging, 10Tracking: Use draft 4 of JSON Schema specification - https://phabricator.wikimedia.org/T46809#4073187 (10Nuria) [16:18:25] 10Analytics: Measure DNT usage across geographies and wikis - https://phabricator.wikimedia.org/T187376#4073188 (10Nuria) [16:18:36] 10Analytics-Kanban, 10Google-Summer-of-Code (2018): Proposal : [Analytics] Improvements to Wikistats2 front-end - https://phabricator.wikimedia.org/T189964#4073189 (10sahil505) @mforns Thanks a lot for reviewing this. I'll make the suggested changes as soon as I can :) [16:19:08] 10Analytics-Kanban: Create canary site for wikistats on labs - https://phabricator.wikimedia.org/T188273#4002016 (10Nuria) [16:19:17] 10Analytics-Kanban: Create canary site for wikistats on labs - https://phabricator.wikimedia.org/T188273#4002016 (10Nuria) 05Open>03Resolved [16:19:53] 10Analytics, 10Analytics-Wikistats: Wikistats 2.0: "aa.wikipedia.org" exists and has data available, but marked "Invalid" - https://phabricator.wikimedia.org/T187414#4073193 (10Nuria) [16:22:23] 10Analytics, 10Analytics-Data-Quality, 10Analytics-Kanban, 10WMDE-Analytics-Engineering, 10User-GoranSMilovanovic: Please review: public data sets for the WDCM Biases Dashboard - https://phabricator.wikimedia.org/T189653#4073199 (10Milimetric) a:03Milimetric [16:26:35] 10Analytics: Provide unqiues estimate/offset breakdowns in AQS - https://phabricator.wikimedia.org/T164593#4073204 (10Nuria) [16:30:40] 10Analytics: Map tooltip and line graph guide are misaligned in Ubuntu Chrome - https://phabricator.wikimedia.org/T187453#4073222 (10Nuria) [16:31:23] 10Analytics, 10Analytics-EventLogging, 10Tracking: Update client-side event validator to support (at least) draft 3 of JSON Schema - https://phabricator.wikimedia.org/T182094#4073225 (10Nuria) [16:32:20] 10Analytics, 10Analytics-Data-Quality, 10Analytics-Kanban, 10WMDE-Analytics-Engineering, 10User-GoranSMilovanovic: Please review: public data sets for the WDCM Biases Dashboard - https://phabricator.wikimedia.org/T189653#4073228 (10Milimetric) +1 from me, @GoranSMilovanovic, this doesn't involve any user... [16:36:27] ping ottomata [16:37:31] hola ottomata [16:38:12] 10Analytics-Kanban: Put in service 8 new hadoop nodes - https://phabricator.wikimedia.org/T182926#4073235 (10elukey) [16:38:18] 10Analytics-Cluster, 10Analytics-Kanban, 10Operations, 10Patch-For-Review, 10User-Elukey: rack/setup/install analytics107[0-7] - https://phabricator.wikimedia.org/T188294#4073238 (10elukey) [16:39:52] ACKKK [16:39:55] haha sorry [16:39:58] deep in these python docs [16:39:59] sheesh [16:40:26] 10Analytics, 10Analytics-Kanban: Write agreggation job for eventlogging page preview data - https://phabricator.wikimedia.org/T188310#4073253 (10Nuria) First task is from eventlogging to parquet data (a la pageview hourly) [16:42:42] 10Analytics, 10Analytics-Kanban: Write agreggation job for eventlogging page preview data - https://phabricator.wikimedia.org/T188310#4073258 (10fdans) a:03mforns [16:42:53] 10Analytics-Kanban: Puppetize job that saves old versions of geoIP database - https://phabricator.wikimedia.org/T136732#4073260 (10fdans) a:03fdans [16:42:59] madhuvishy: hiiiii, not sure how familiar you are with venv site packages jupyter systemd spawner stuff, but if at all, would love to ask you some qs ina bit (after our meetings are over) [16:44:38] ottomata: i'm not sure either, but busy day until around 1pm here, chat after? if not tomorrow morning is good :) [16:49:37] k [16:51:23] 10Analytics, 10EventBus, 10Services (watching): Failure in EventBus schema for mediawiki/revision/visibility-change - https://phabricator.wikimedia.org/T187362#4073312 (10fdans) [16:52:33] 10Analytics, 10Discovery, 10Wikidata, 10Wikidata-Query-Service, 10Wikimedia-Stream: Increase kafka event retention to 14 or 21 days - https://phabricator.wikimedia.org/T187296#4073313 (10fdans) [16:54:04] 10Analytics, 10Analytics-Kanban: Write agreggation job for eventlogging page preview data - https://phabricator.wikimedia.org/T188310#4073337 (10fdans) [16:54:06] 10Analytics-Kanban, 10MW-1.31-release-notes (WMF-deploy-2018-02-27 (1.31.0-wmf.23)), 10Patch-For-Review: Record and aggregate page previews - https://phabricator.wikimedia.org/T186728#4073335 (10fdans) [16:54:37] 10Analytics-Kanban, 10MW-1.31-release-notes (WMF-deploy-2018-02-27 (1.31.0-wmf.23)), 10Patch-For-Review: Record and aggregate page previews - https://phabricator.wikimedia.org/T186728#3953218 (10fdans) a:03mforns [16:55:35] 10Analytics, 10Analytics-Wikistats: Wikistats 2: New Pages split by editor type wrongly claims no anonymous users create pages - https://phabricator.wikimedia.org/T185342#4073350 (10fdans) [16:55:58] 10Analytics, 10Operations, 10User-Elukey: Tune Varnishkafka delivery errors to be more sensitive - https://phabricator.wikimedia.org/T173492#4073351 (10fdans) [16:56:45] 10Analytics, 10Analytics-Cluster, 10Patch-For-Review: Move EventStreams to new jumbo cluster. - https://phabricator.wikimedia.org/T185225#4073354 (10fdans) [16:58:45] 10Analytics: Regular backups for EL data - https://phabricator.wikimedia.org/T183383#4073360 (10fdans) [16:59:08] 10Analytics: Regular backups for EL data - https://phabricator.wikimedia.org/T183383#3851649 (10fdans) 05Open>03declined [16:59:59] 10Analytics, 10User-Elukey: Refactor analytics cronjobs to alarm on failure with (maybe) cronic - https://phabricator.wikimedia.org/T172532#4073367 (10fdans) [17:00:14] hey joal [17:01:01] 10Analytics-Kanban, 10EventBus, 10Services (next): Malformed HTTP message in EventBus logs - https://phabricator.wikimedia.org/T178983#4073379 (10Ottomata) 05Open>03Resolved I think we can close this. Reopen if not. [17:01:52] 10Analytics, 10EventBus, 10Services (next): Malformed HTTP message in EventBus logs - https://phabricator.wikimedia.org/T178983#4073381 (10fdans) [17:02:09] 10Analytics, 10Analytics-Kanban, 10Analytics-Wikistats: Wikistats Bug – Put view settings in URL so it can be shared - https://phabricator.wikimedia.org/T179444#4073382 (10Milimetric) [17:02:27] 10Analytics, 10Analytics-Kanban, 10Analytics-Wikistats: Consistently preserve settings when a user switches to a new metric (especially on the same page). - https://phabricator.wikimedia.org/T183181#4073383 (10Milimetric) [17:04:00] 10Analytics: Fix wikimedia-history revision-deleted data - https://phabricator.wikimedia.org/T178587#4073386 (10fdans) [17:05:57] 10Analytics, 10User-Elukey: Alarms on pageview API latency increase - https://phabricator.wikimedia.org/T164243#4073387 (10fdans) [17:07:05] 10Analytics, 10Wikimedia-Stream, 10Wikimedia-Incident: Alerts for common/important EventStreams topic volume - https://phabricator.wikimedia.org/T174493#4073389 (10fdans) [17:10:31] 10Analytics: Fix wikimedia-history revision-deleted data - https://phabricator.wikimedia.org/T178587#4073398 (10JAllemandou) [17:30:31] 10Analytics, 10Analytics-EventLogging, 10Analytics-Kanban, 10User-Elukey: Run eventlogging purging script on beta labs to avoid disk getting full - https://phabricator.wikimedia.org/T171203#4073466 (10Nuria) 05Open>03Resolved [17:39:31] 10Analytics-Kanban, 10Patch-For-Review: Wikistats: labeling of pageviews is wrong on table and graph. Issues with dates - https://phabricator.wikimedia.org/T189266#4073502 (10Nuria) 05Open>03Resolved [17:41:46] varnent: we have fixed the issues you reported with labelling of pageviews on wikistats, data that was for february was seen by some users with the "january" label . Please take a look: http://localhost:5000/dist-dev/#/es.wikipedia.org/reading/total-pageviews [17:45:40] milimetric: i am going to schedule a meeting with asaf for 1st week of april, does that sound ok? [17:46:05] sounds good nuria_ [17:52:35] milimetric: ok, done [17:55:05] nuria_: btw, i pasted the whereshow stuff in reference to data map [17:55:05] https://github.com/linkedin/WhereHows/blob/master/wherehows-docs/use-cases.md#use-cases [17:55:08] might help, who knows [17:55:12] also would be a buncha work to set up [17:56:42] ottomata: wow [17:56:47] ottomata: i totally missed this [17:56:59] ottomata: seems that it would be huge to maintain [17:57:16] ya [17:57:23] i mean, it does a lot of auto discovery in hdfs, etc. [17:57:32] and then presents the stuff it finds in a a gui/db [17:57:38] and let's you augment it with metadata [17:58:02] i think also has stuff for comliance, governence, etc. [17:59:40] i think it just crawls for data and schemas and let's you view and manage them in a nice web app [18:02:05] 10Analytics, 10Analytics-Kanban: Issues with page deleted dates on data lake - https://phabricator.wikimedia.org/T190434#4073560 (10Nuria) [18:02:17] ottomata: I made some little changes to https://gerrit.wikimedia.org/r/#/c/415218/ (performance kafka coal) and ready to be deployed [18:02:53] I'd merge it (with lmarlier support) if you are ok [18:03:02] 10Analytics, 10Analytics-Kanban: Issues with page deleted dates on data lake - https://phabricator.wikimedia.org/T190434#4073572 (10Nuria) Ping @Nettrom did you ended up filing a ticket for the issue described? I could not find it. (cc @joal) If not : could you describe issue a bit more? [18:03:36] Nettrom: hello! see if this tickets rings a bell: https://phabricator.wikimedia.org/T190434#4073572 [18:03:55] (Imarlier sorry :) [18:03:57] nuria_: just saw the notification popup, I’ve got some time available and will start digging into this today :) [18:04:12] joal and /me where trying to look for ticket on the matter but could not find original ticket, Thanks! [18:04:35] I never got around to filing a ticket because I wanted to make sure I had some data [18:05:01] never had time to gather the data because ACTRIAL grabbed all of it. now that I’ve got time, I’ll get some data and dig into it [18:05:17] reading [18:07:13] elukey: am confused, do they want to have two different hiera settings for coal_whisper_dir? [18:07:22] also, do they ever vary it anywhere (labs, etc.?) [18:07:36] if not, maybe just refering to /var/lib/coal is fine? [18:08:02] 10Analytics, 10CirrusSearch, 10Discovery, 10Discovery-Search, and 3 others: Expose a metric that reflect EventBus queue pressure - https://phabricator.wikimedia.org/T190416#4073581 (10mobrovac) We have these metrics already exposed in Grafana: - [backlog in number of messages](https://grafana.wikimedia.org... [18:09:41] ottomata: I put those since they were originally referring to a heira/common/performance/coal key, that was set to /var/lib/coal [18:09:44] 10Analytics, 10Analytics-Wikistats: Beta Release: Remaining UI advice from Erik - https://phabricator.wikimedia.org/T182109#3813007 (10mforns) Looking at this task, all of its issues are already solved by other tasks, except for: > https://meta.wikimedia.org/wiki/Research:Page_view > That page is almost usele... [18:09:59] ottomata: it is a way to vary it if needed, should be fine with those defaults [18:10:06] elukey: ya, but there are two settings now [18:10:25] ::coal::coal_whsiper_dir and ::site::coal_whisper_dir [18:10:28] maybe shoudl just have one? [18:10:40] have the site one reference the ::coal one? [18:11:26] could be an option, but I don't see the problem of having two hiera settings [18:11:35] i think they want them to always be set to the same thing [18:12:12] does the performance::site have to be run on the same host? [18:12:17] it seems so, since it symlinks to /var/lib/coal [18:12:25] so, i'd make a dependency [18:12:41] nuria_: Woot! Thank you!! [18:12:44] and then reference the puppet var, e.g. $::profile::performance::coal::coal_whisper_dir [18:12:51] and not have a param on the ::site class [18:13:09] Do you have any sense of when project family stat\s will be in v2? :) [18:13:34] 10Analytics, 10CirrusSearch, 10Discovery, 10Discovery-Search, and 3 others: Expose a metric that reflect EventBus queue pressure - https://phabricator.wikimedia.org/T190416#4073598 (10dcausse) I think that per queue is perfectly OK for my usecase. Can we make these numbers accessible from MediaWiki? [18:13:35] ottomata: ok can do it, lemme amend it (otherwise we'll keep discussing until tomorrow :) [18:13:56] k [18:16:11] varnent: we will not be working on that on the UI until probably 6 months from now, data will probably be available in the APIs but not on the UI for a bit (cc milimetric joal fdans mforns ) [18:16:33] ottomata: better https://gerrit.wikimedia.org/r/#/c/415218/ ? [18:18:41] Let's do it elukey! [18:18:51] i think maybe ian is not around? [18:18:54] shoudl we wait? [18:19:33] he is! Waiting for our +2 [18:19:48] going to ping him and then merge :) [18:19:51] thanks for the review! [18:19:55] ok cool [18:19:56] yup [18:20:26] 10Analytics: Correct mixed case in Wikistats2 - https://phabricator.wikimedia.org/T190435#4073621 (10mforns) [18:20:51] 10Analytics, 10Analytics-Wikistats: Correct mixed case in Wikistats2 - https://phabricator.wikimedia.org/T190435#4073621 (10mforns) [18:24:43] 10Analytics, 10Analytics-Wikistats: Correct mixed case in Wikistats2 - https://phabricator.wikimedia.org/T190435#4073654 (10Amitjoki) a:03Amitjoki [18:26:09] nuria_: Hmmm..okay - Comms uses those stats with press - should we set up something for ourselves somewhere or will stats in v1 be reliable for awhile longer? [18:26:55] nuria_: We try to update these every 2-3 months: https://meta.wikimedia.org/wiki/Communications/Communicating_about_the_Wikimedia_Foundation#Statistics [18:28:26] varnent: i think wikistats 1 will have that information for the next year most likely, remember that wikistats is made for the community rather than WMF and thus they have a say when it comes to prioritization of what do we work on next [18:30:18] nuria_: Yes - although these stats are also used by many within the community that work with press. Part of the reason we posted them on Meta-Wiki was we got so many requests from community for them. :) [18:30:37] varnent: our next steps in wikistats for the upcoming quarter have a lot to do with UI, next we will work on backend again (which means that although that data might be computed the ui will not have it) and later back to the ui , i am kind of guessing here but 6 months seems about right [18:30:53] Cool beans - thank you! :) [18:31:19] varnent: ya, we do have those requests too, they are in phabricator. they are just lower priority than computing say, "editors" for ALL projects [18:31:32] varnent: which we currently do not do yet, if that makes sese [18:36:34] 10Analytics, 10Analytics-Wikistats: Beta Release: Remaining UI advice from Erik - https://phabricator.wikimedia.org/T182109#4073694 (10mforns) This is the task I created T190435. Will rename and modify the description of this one. [18:39:52] 10Analytics, 10Analytics-Wikistats: Present a page view metric description to the user that they are likely to understand - https://phabricator.wikimedia.org/T182109#4073707 (10mforns) [18:43:25] all right people, logging off! byyeee [18:43:55] byee [18:47:11] 10Analytics, 10Analytics-Kanban, 10Analytics-Wikistats: Wikistats Beta – Put view settings in URL so it can be shared. bookmarks - https://phabricator.wikimedia.org/T179444#4073762 (10Nuria) [18:49:13] 10Analytics-Kanban, 10Analytics-Wikistats: Beta Release: Wikistats: support annotations in graphs - https://phabricator.wikimedia.org/T178015#4073779 (10Nuria) [18:51:33] 10Analytics, 10Analytics-Wikistats: Intervals/buckets for data arround pageviews per country in wikistats maps - https://phabricator.wikimedia.org/T188928#4073786 (10Nuria) [19:03:54] 10Analytics, 10Analytics-Wikistats: Wikistats 2.0: "aa.wikipedia.org" exists and has data available, but marked "Invalid" - https://phabricator.wikimedia.org/T187414#4073814 (10Krinkle) @fdans Where is this list maintained, and what is the process for updating it? Given this is not a new wiki, it suggests that... [19:04:24] madhuvishy: nm, i think i figured it out [19:04:51] don't exactly know why, but if i want to use system-site-packages in the notebook venv, i also need to install jupyter etc. into the venv with --ignore-installed [19:05:06] now i have best of both worlds! system packages work, but venv packages take precedence [19:05:32] 10Analytics-Kanban, 10Patch-For-Review: Refresh SWAP notebook hardware - https://phabricator.wikimedia.org/T183145#4073817 (10Ottomata) Ah, ok, a better user venv upgrade is: ``` wheels_path=/srv/jupyterhub/deploy/artifacts/stretch/wheels for u in $(getent passwd | awk -F ':' '{print $1}'); do venv=/home/... [19:05:38] 10Analytics-Kanban, 10Patch-For-Review: Refresh SWAP notebook hardware - https://phabricator.wikimedia.org/T183145#4073818 (10Ottomata) I've added https://wikitech.wikimedia.org/wiki/SWAP#Administration for future work [19:06:15] ok, so joal, i can't really update any dependencies users have installed for python 3.4 [19:06:18] they'll have to do that themselves [19:06:32] on the other hand, i installed both pyhive and impyla for all users :) [19:06:41] and also tons of system site packages, pandas, scipi, etc. [19:08:32] 10Analytics: Spark notebook integration - https://phabricator.wikimedia.org/T190443#4073838 (10Ottomata) [19:34:31] 10Analytics, 10ChangeProp, 10EventBus, 10MediaWiki-JobQueue, and 3 others: [EPIC] Develop a JobQueue backend based on EventBus - https://phabricator.wikimedia.org/T157088#4073896 (10Pchelolo) [19:34:36] 10Analytics, 10ChangeProp, 10EventBus, 10MediaWiki-JobQueue, and 2 others: Set up grafana alerts for JobQueue-EventBus - https://phabricator.wikimedia.org/T189038#4073893 (10Pchelolo) 05Open>03Resolved We now have a couple of alerts - it should fire when the backlog is too high or when a rule suddenly... [19:35:30] 10Analytics-Kanban, 10Patch-For-Review: Refresh SWAP notebook hardware - https://phabricator.wikimedia.org/T183145#4073897 (10Ottomata) Email sent (Subject: 'New SWAP (Jupyter Notebook) servers and updates!'). Timeline for notebook1001 deprecation: Monday April 2nd. [19:46:39] That's already great ottomata ) [19:47:13] nuria_: Thanks for filing the ticket from conversation with Nettrom :) Nettrom -- Thanks for helping us providing better data ! [19:47:43] dsaez: Heya - still here? [19:48:39] joal: k [19:49:21] nuria_: no wonder I couldn't find it in phab nor email - I didn't remember it had happened in IRC ... Thanks :) [19:49:45] joal: somehow i think i dreamed there was a ticket [19:49:49] ... [19:49:53] so did i nuria_ ! [20:12:46] bearloga: tell me about R kernel [20:24:34] ottomata: yo. I'm not even sure when the problem is occurring. did you follow the steps at https://irkernel.github.io/installation/ ? I wonder if madhuvishy [20:24:49] had to do anything special when she got it working on notebook1001 [20:39:38] bearloga: my R experience is next to Nil [20:39:38] :) [20:40:27] i don't know of others that use it much, i assume you were able to install it in your own homeidr somehow? [20:42:11] trying [20:42:14] trying this https://irkernel.github.io/docs/IRkernel/0.7/ [20:47:17] ottomata: madhu installed the R kernel on the machine for all users (for my & Chelsy's use). I haven't tried installing it myself. I can try. I'm worried about permissions at the final step `IRkernel::installspec()` [20:48:00] bearloga: i think it will work! [20:48:06] it installs to a local user prefix [20:48:08] i just did it [20:48:13] i was able to start the R kernel [20:48:17] not sure what to do once i'm there [20:48:23] but ya, we hsould have kernel's like this installed for all users [20:48:35] but, following those instructions seemed to work for me [20:48:45] > IRkernel::installspec() [20:48:45] [InstallKernelSpec] Installed kernelspec ir in /srv/home/otto/.local/share/jupyter/kernels/ir [20:48:49] by the way, "To install system-wide, set user to False in the installspec command: `IRkernel::installspec(user = FALSE)` [20:49:00] it seems that was default! [20:50:34] 10Analytics: Support R Kernels by default for all users. - https://phabricator.wikimedia.org/T190453#4074109 (10Ottomata) [20:50:38] bearloga: ^ :) [20:51:19] ottomata: +1 +1 +1 :D [20:53:35] chelsyx: idk if you did this already but I'd recommend running `update.packages(ask = FALSE, checkBuilt = TRUE)` on the new notebook machine because otherwise you might encounter issues using packages that were/are linked to external libraries [20:54:35] bearloga: Thanks for reminding! [20:55:30] ottomata: I'm going to try installing the kernel for myself and let you know how that goes. but yes, thumbs up to the ticket to make it available system-wide :) [21:00:54] joal: yt? [21:03:52] yup nuria_ [21:04:24] joal: for the mobile jobs there are tow tables [21:05:09] joal: mobile_apps_session_metrics and mobile_apps_session_metrics_by_os [21:05:17] right [21:05:45] joal: mobile_apps_session_metrics does not have any data for march [21:06:39] joal: wait is that data monthly but every week? what? i think i am getting confused [21:06:56] nuria_: oozie bundle definition stats : period-duration for global is 30 days, period duration by os is 7 days [21:07:11] nuria_: Don't ask me why :) [21:07:25] joal: the metrics by os are weekly then [21:07:47] joal: but the metrics by global are monthly but calculated weekly? [21:09:25] 10Analytics: Support R Kernels by default for all users. - https://phabricator.wikimedia.org/T190453#4074109 (10mpopov) Mentioned this on IRC already but also mentioning it here :) In https://irkernel.github.io/installation/ it says: >To install system-wide, set user to False in the `installspec` command: `IRke... [21:11:54] nuria_: From what I read in oozie for global, to compute feb25 data, the system needs (feb25+30 days) -- Therefore not present yet [21:12:14] joal: right, so it calculates every week monthly data [21:12:32] correct nuria - Data is computed weekly (freq doesn't change) [21:12:44] joal: boy that seems not the best idea given how we calculate everything else [21:12:44] However, dependency are different (7 days or 30 days) [21:13:30] chelsyx: do you have time for 1 question? [21:13:49] nuria_: yes [21:14:28] chelsyx: do we ever use the data for mobile_apps_session_metrics w/o os component? that is the job that agreggates android and IOS [21:14:29] ottomata: everything's good re: installing r kernel in my homedir :) [21:15:33] great [21:15:34] nuria_: Not that I know of. Should check with HaeB . [21:16:33] chelsyx: the way those jobs are calculated is real strange , i am going to open a ticket to stop calculating the "mobile_apps_session_metrics" that is not split by OS [21:17:45] nuria_: recall https://phabricator.wikimedia.org/T117615 [21:18:46] i honestly don't know why it was calculated in the aggregate in the first place, which is why i filed that task back then to separate it into ios and android [21:18:47] HaeB: meaning that that job should have died ages ago? [21:19:05] that's what it sounds like since we have mobile_apps_session_metrics_by_os [21:19:07] HaeB: ya, agreed, seems not very useful. Ok, will be killing that job [21:19:08] as noted there, we kept it around for a while to assess historical trends [21:19:12] cc chelsyx joal [21:19:18] but by now we can stop that, yes [21:20:09] 10Analytics: Phasing away one of the mobile apps session metrics jobs. - https://phabricator.wikimedia.org/T190459#4074265 (10Nuria) [21:20:44] joal: I am going to stop this one: https://hue.wikimedia.org/oozie/list_oozie_coordinator/0005625-180308085316299-oozie-oozi-C/ [21:21:03] joal: and submit a patch that removes code, sounds good? [21:21:08] nuria_: no prob for me [21:21:36] nuria_: no prob for me. can't think of anything that will use that data [21:21:47] nuria_: patch will be about removing bundle and transform it to a corrd (since only one of the thing runs now) [21:22:39] 10Analytics: Phasing away one of the mobile apps session metrics jobs. - https://phabricator.wikimedia.org/T190459#4074265 (10Tbayer) Please update the documentation at https://wikitech.wikimedia.org/wiki/Analytics/Data_Lake/Traffic/mobile_apps_session_metrics once the generation of new data for `mobile_apps_ses... [21:28:08] 10Analytics: Phasing away one of the mobile apps session metrics jobs. - https://phabricator.wikimedia.org/T190459#4074302 (10Nuria) Job is been killed on https://hue.wikimedia.org/oozie/list_oozie_coordinator/0005625-180308085316299-oozie-oozi-C/?bundle_job_id=0005624-180308085316299-oozie-oozi-B [21:33:18] HaeB: we will be dropping that data once code is removed right? [21:33:34] no, i think we should keep it [21:33:49] like i said, it has some limited value in assessing historical trends [21:59:15] (03CR) 10Nuria: "Let's please consolidate the tickets we have in kanban about this if this patch hits more that one ticket. Or, rather, split changes per t" [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/416999 (https://phabricator.wikimedia.org/T187345) (owner: 10Fdans) [22:02:11] 10Analytics-Kanban: The size of metric areas in the dashboard should scale to available window space - https://phabricator.wikimedia.org/T187345#3972459 (10Nuria) {F15975378} You probably know this but FYI that this changeset looks like this on desktop so it cannot be deployed as is. [22:08:14] ottomata: Had q quick loof at hoodie - Seems super interesting :) [22:23:20] 10Analytics, 10EventBus, 10MediaWiki-JobQueue, 10Goal, 10Services (doing): FY17/18 Q3 Program 8 Services Goal: Migrate two high-traffic jobs over to EventBus - https://phabricator.wikimedia.org/T183744#4074461 (10mobrovac) [22:23:23] 10Analytics, 10ChangeProp, 10EventBus, 10MediaWiki-JobQueue, and 3 others: [EPIC] Develop a JobQueue backend based on EventBus - https://phabricator.wikimedia.org/T157088#4074462 (10mobrovac) [22:23:29] 10Analytics, 10ChangeProp, 10EventBus, 10MediaWiki-JobQueue, and 2 others: Set up grafana alerts for JobQueue-EventBus - https://phabricator.wikimedia.org/T189038#4074460 (10mobrovac) [22:45:06] (03CR) 10Nuria: "Looks good overall although cannot speak for CSS, comments refer to how to have a global property that refers to display" (034 comments) [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/416999 (https://phabricator.wikimedia.org/T187345) (owner: 10Fdans) [22:51:54] (03CR) 10Nuria: "This change does not render for me on desktop or mobile, dashboard is not visible." [analytics/wikistats2] - 10https://gerrit.wikimedia.org/r/421001 (https://phabricator.wikimedia.org/T187440) (owner: 10Fdans) [22:55:17] 10Analytics, 10Analytics-Kanban: Issues with page deleted dates on data lake - https://phabricator.wikimedia.org/T190434#4074516 (10Nettrom) As mentioned on IRC earlier today, I never filed a ticket because I didn't have the time to sit down and make sure I had data that allowed me to understand exactly what t... [23:03:39] 10Analytics, 10Analytics-Kanban: Issues with page deleted dates on data lake - https://phabricator.wikimedia.org/T190434#4074529 (10Nuria) Sounds awesome, do let us know of your findings. [23:21:27] nuria_: Do you have a minute? [23:55:44] someone broke yarn :( it's getting 404 for resources such as https://yarn.wikimedia.org/static/jquery/jquery-1.8.2.min.js [23:56:31] yarn webui i mean