[00:12:51] Nicely targeted :) [00:13:39] I’m just happy that I’m finally aware of the difference between machines, thus will do my queries on the mariadb-only host rather than wasting hadoop resources. [00:18:27] At least you have access to Hadoop resources [00:26:49] lol. right? considering that this has only happened once, I’m pretty happy with stat machines [03:25:16] 10Analytics-Kanban, 10RESTBase-API, 10Services (later), 10User-mobrovac: Expose pageview data in each project's REST API - https://phabricator.wikimedia.org/T119094#3488352 (10mobrovac) [04:25:27] 10Analytics-Kanban, 10RESTBase-API, 10Services (later), 10User-mobrovac: Expose pageview data in each project's REST API - https://phabricator.wikimedia.org/T119094#3488437 (10Nuria) 05Open>03declined [04:46:54] 10Analytics-Kanban, 10RESTBase-API, 10Services (later), 10User-mobrovac: Expose pageview data in each project's REST API - https://phabricator.wikimedia.org/T119094#3488443 (10Nuria) Declining cause we do not feel this items delivers enough value to change our current url scheme, also semantics proposed br... [07:03:37] !log restarted mobile_apps-session_metrics-coord-global-30days failed job via Hue [07:03:38] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [07:58:14] !log suspended webrequest-load-bundle as prep step to restart the hive daemons [07:58:15] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [07:58:29] now I need to figure out when to restart them, too many people running queries atm :) [07:59:42] RoanKattouw,awight - thanks for the notes! I can't reproduce the issue at the moment, maybe it was temporarily? We are running heavy update queries on dbstore1002 to sanitize data, it might be the issue [09:14:15] elukey: The db is playing nicely for me again, but I’m happy to report back with any more turbulence I encounter in the future! [09:16:10] awight: yes please! I am pretty sure it is the current cleaning script, sorry for the trouble! [09:32:26] Sounds like a fun job… I might be doing something similar, munging two different formats of log_params :) [09:41:00] if you are curious you can check the eventlogging_cleaner.py script in operations/puppet :) [09:41:37] basically we are sanitizing tables in the log database, either dropping rows with timestamp > 90 days or setting to NULL the sensitive fields [09:41:49] (to keep history) [09:43:09] oh nice, thanks for the privacy upgrade!! [10:05:02] !log suspended again webrequest-load-bundle as prep step to restart the hive daemons [10:05:03] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [10:18:01] Hi A-Team! [10:20:23] * elukey waves to addshore [10:20:42] Got a question I guess most people in the A-Team will know :D [10:20:54] shoot! [10:21:08] I'm going to answer https://phabricator.wikimedia.org/T172100, and basically want to know if the data will need to be hidden to only people with NDAs or if public is okay [10:21:34] Essentially the data is number of calls to the api with action=wbeditentity&clear= mapped to user agents [10:22:34] when it comes to mapping reqs to UA I always suggest to use NDA to be sure [10:22:43] okay! :) [10:23:07] and I understand the best way to do that on phab is make a paste that is only accessible to NDA people, and then you can embed that in the ticket, cool! :) [10:23:15] addshore: I saw that you are running some queries with hive, I need to restart the hive daemons so if you see it killed it might be me :( [10:23:19] will try to wait for its completion [10:23:27] addshore: +1 [10:23:48] okay! yeh, this query shouldnt take too much longer! I can ping you when its done!!!!! [10:24:05] It's for the ticket linked above ;) [10:25:15] super :) [10:27:03] elukey: done! [10:32:17] 10Analytics-Cluster, 10Analytics-Kanban, 10Patch-For-Review, 10User-Elukey: Understand Kafka ACLs and figure out what ACLs we want for production topics - https://phabricator.wikimedia.org/T167304#3488802 (10elukey) This is the result of adding ACLs for user `test` to produce/consume to the `elukey2` topic... [10:40:03] 10Analytics-Cluster, 10Analytics-Kanban, 10Patch-For-Review, 10User-Elukey: Understand Kafka ACLs and figure out what ACLs we want for production topics - https://phabricator.wikimedia.org/T167304#3488833 (10elukey) As explained before we also need to explicitly set ACLs for cluster operations between brok... [10:51:37] * elukey lunch! [11:37:37] ok so I restarted again webrequest-load-bundle, I forgot that today is the first of the month and uniques big jobs are running [11:37:40] will try again later on [11:50:07] 10Analytics, 10EventBus, 10Operations, 10User-Elukey: Eventbus does not handle gracefully changes in DNS recursors - https://phabricator.wikimedia.org/T171048#3452124 (10elukey) p:05High>03Low The remaining step is to explore the possibility of having a logic to cache the statsd IP only for a limited a... [12:14:07] 10Analytics-EventLogging, 10Analytics-Kanban, 10Community-Tech, 10DBA, 10User-Elukey: Drop CookieBlock* tables from EventLogging DB - https://phabricator.wikimedia.org/T171883#3488967 (10elukey) Action executed on db1046 (m4-master): ``` MariaDB [(none)]> use log; Reading table information for completio... [12:39:36] 10Analytics-Cluster, 10Analytics-Kanban, 10User-Elukey: Perf test RAID vs JBOD with new hardware and kafka versions - https://phabricator.wikimedia.org/T168538#3488980 (10elukey) I had an interesting chat with the Ops team about this task and I believe that we don't need to spend ton of time working on this... [12:45:30] elukey: ok, what's up with geowiki? [12:46:05] milimetric: o/ [12:46:12] howdy [12:46:52] nothing big, I noticed the cronspam alert and tried to check what was the weirdness.. two days ago I thought that the issue was only the rsync target (was stat1002 now is stat1006) [12:47:07] so the git repo might just need to be reset [12:47:19] isn't 1005 the new 1002? [12:47:25] yes sorry [12:47:31] lemme check [12:47:37] (03CR) 10Zhuyifei1999: [C: 032] Set session.permanent = True when user is logged in [analytics/quarry/web] - 10https://gerrit.wikimedia.org/r/368745 (https://phabricator.wikimedia.org/T164390) (owner: 10Zhuyifei1999) [12:47:57] (03Merged) 10jenkins-bot: Set session.permanent = True when user is logged in [analytics/quarry/web] - 10https://gerrit.wikimedia.org/r/368745 (https://phabricator.wikimedia.org/T164390) (owner: 10Zhuyifei1999) [12:48:01] milimetric: stat1006 now hosts the geowiki bare repo though [12:48:08] so thorium syncs from it [12:48:13] hm, but geowiki was running on 1003 I think [12:48:22] right ok [12:48:46] so yeah I confused stat1002/3 sorry :) [12:49:16] !log restart hive daemons on analytics1003 to pick up new jvm settings (bigger Xmx, JMX ports) [12:49:17] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [12:49:50] ok, then I'll do the stash thing I was saying on email [12:49:55] and re-run, see what's up [12:50:55] milimetric: what do you mean re-run? (asking because I am ignorant about geowiki) [12:51:53] well geowiki runs in that cron, the command's there [12:52:08] so it's fine to rerun it, should do nothing if it already finished successfully [12:52:24] but for whatever reason it seems in a bad state [12:52:28] I thought that the command was only a check, and puppet did the git pull [12:52:44] (from the bare repo that in this case is on the same host) [12:52:49] we might be talking about different things, let me check to see what's in the cron [12:53:26] I'm talking about # Puppet Name: geowiki-process-data [12:53:40] that's what crunches the data, and updates the repository [12:53:46] ahh okok [12:54:31] it might be one of the other commands though, 'cause I tested the process data and that seemed fine [12:56:20] milimetric: so the cron creates new things in the private repo, then it pushes them to the bare one? [12:56:33] (that is in turn copied over to thorium) [12:56:58] I don't remember exactly, I think there are a few more steps, but there are at least those three you mentioned [12:57:03] crunch - move - sync [12:57:07] 10Quarry, 10Patch-For-Review: Quarry should remember my login - https://phabricator.wikimedia.org/T164390#3489031 (10zhuyifei1999) 05Open>03Resolved a:03zhuyifei1999 It should now remember login for 31 days (default). If this is too long or too short, or if a "remember my login" checkbox should be added,... [12:58:16] milimetric: thanks! [13:00:50] there were some errors when we first moved the job from stat1006, because the versions of python-mysql client were different, so I'm thinking this might be related [13:06:02] ok, elukey the process_data part was really quick and didn't leave the repo in a bad state. [13:06:30] I'm running the "make_and_push_limn_files" part which takes at least an hour [13:06:42] it's in a screen on stat1006 [13:07:01] (and it's in verbose mode so we can figure out if anything goes wrong) [13:08:50] super [13:10:04] 10Analytics-Kanban, 10Discovery, 10Discovery-Analysis, 10Patch-For-Review: Add purge info for Kartographer schema - https://phabricator.wikimedia.org/T171622#3489068 (10mforns) I created a Gerrit change to add Kartographer to EL white-list. Please review and +1 if it looks good, we'll then merge it. Thanks! [13:11:03] 10Analytics-Cluster, 10Analytics-Kanban, 10Patch-For-Review: Replacement of stat1002 and stat1003 - https://phabricator.wikimedia.org/T152712#3489069 (10Ottomata) @Catrope just emailed: > I would love to migrate to stat1006 from stat1003, but stat1006 is unusably slow right now while stat1003 is snappy. Con... [13:12:44] 10Analytics-Cluster, 10Analytics-Kanban, 10Patch-For-Review: Replacement of stat1002 and stat1003 - https://phabricator.wikimedia.org/T152712#3489090 (10Ottomata) I can't seem to reproduce... [13:15:42] 10Analytics-Cluster, 10Analytics-Kanban, 10User-Elukey: Perf test RAID vs JBOD with new hardware and kafka versions - https://phabricator.wikimedia.org/T168538#3489094 (10Ottomata) +1 :) [13:27:14] 10Analytics-Kanban: Practice with photorec - https://phabricator.wikimedia.org/T171972#3489111 (10Ottomata) [13:34:00] 10Analytics-Kanban: Practice with photorec - https://phabricator.wikimedia.org/T171972#3489117 (10Ottomata) I did a little test on stat1003. Ubuntu/Debian has a package called 'testdisk' that ships with photorec! :) I made an 10G LVM XFS partition on stat1003, copied in some files, including one sampled gz log... [13:41:08] 10Analytics-Cluster, 10Analytics-Kanban, 10Operations, 10User-Elukey: thorium - failed git clone of geowiki-data-private - https://phabricator.wikimedia.org/T171923#3489126 (10Ottomata) Weird, I'm not sure what's up here. The /var/lib/stats/.gitconfig file looks good, and it is [[ https://github.com/wikim... [13:49:27] one comment elukey [13:50:56] ottomata: exactly what I wanted to discuss with you :) [13:51:39] let's use the Hive namespace then [13:52:51] updated code review [13:53:45] so in graphite will be: Hive -> Server/Metastore -> hostname -> Hive -> Server/Metastore -> JvmMetrics [13:53:51] like the hadoop daemons [13:55:46] +1 [13:57:58] there is also our dear oozie that is without jmx ports [14:02:45] aye :) [14:02:53] hey a-team, i need some room in labs [14:02:57] do we need [14:03:14] https://horizon.wikimedia.org/project/instances/5f0cc6a2-d20d-4e1d-9377-340de146d719/ ?https://horizon.wikimedia.org/project/instances/db12108f-870d-4386-bf93-3eaf63461c99/ [14:03:14] ? https://horizon.wikimedia.org/project/instances/381578bf-d3ac-4f3f-bd7f-54fdcf995350/ [14:03:14] ? https://horizon.wikimedia.org/project/instances/6dcea35d-7b7d-451f-9828-4102e9656286/ [14:03:14] ? [14:03:16] ooops [14:03:29] paws-internal-01, wiki-talk-test, maintenance, elastic1, shiny [14:03:32] do we need those or can I delete? [14:03:47] paws internal ask madhuvishy [14:03:57] ottomata, I don't use any of those [14:04:00] I don't know what others are [14:04:28] I don't use any of those but I'd feel sad to kill 'shiny' [14:04:45] i wonder if it hosting oliver dashboards [14:06:04] this [14:06:04] http://datavis.wmflabs.org/ [14:08:57] ok, i'm deleting wiki-talk-test and maintenance and elastic1 [14:09:00] :) [14:09:03] objections? [14:09:09] orrr i guess i can wait for standup... [14:09:09] :) [14:14:04] kill them :) [14:47:58] 10Analytics-Kanban, 10RESTBase-API, 10Services (later), 10User-mobrovac: Expose pageview data in each project's REST API - https://phabricator.wikimedia.org/T119094#3489358 (10mobrovac) 05declined>03Open >>! In T119094#3488443, @Nuria wrote: > Declining cause we do not feel this items delivers enough v... [15:00:26] 10Analytics-Kanban, 10RESTBase-API, 10Services (later), 10User-mobrovac: Expose pageview data in each project's REST API - https://phabricator.wikimedia.org/T119094#3489446 (10Nuria) >having pageview data appear there (a) completes the set; and (b) informs the user of the availability of the data for the p... [15:03:17] 10Analytics-Kanban: Practice with photorec - https://phabricator.wikimedia.org/T171972#3489458 (10Nuria) [15:07:12] 10Analytics-Kanban, 10User-Elukey: Upgrade AQS to node 6.11 - https://phabricator.wikimedia.org/T170790#3489467 (10Nuria) [15:36:40] 10Analytics-Kanban, 10Patch-For-Review, 10User-Elukey: Wrong JVM heap size set for Hive* daemons - https://phabricator.wikimedia.org/T172107#3489548 (10elukey) [15:36:47] 10Analytics-Cluster, 10Analytics-Kanban, 10User-Elukey: Perf test RAID vs JBOD with new hardware and kafka versions - https://phabricator.wikimedia.org/T168538#3489549 (10elukey) [15:40:44] 10Analytics-Kanban: Upgrade Druid to 0.9.2 as a temporary measure - https://phabricator.wikimedia.org/T170590#3489554 (10Nuria) [15:46:28] 10Analytics-Kanban, 10User-Elukey: Archive PageContentSaveComplete in hdfs while we continue collecting data - https://phabricator.wikimedia.org/T170720#3489572 (10elukey) p:05Triage>03High [15:53:59] 10Analytics-Kanban, 10User-Elukey: dbstore1002 /srv filling up - https://phabricator.wikimedia.org/T168303#3489601 (10elukey) In https://phabricator.wikimedia.org/T170720 we are planning to move one big table (~500GB) to HDFS and drop it from mysql. [15:55:52] * mforns commutes back home (back in 45 mins) [15:56:37] 10Analytics-Kanban, 10Analytics-Wikistats, 10Continuous-Integration-Infrastructure, 10Release-Engineering-Team (Kanban): Fix Wikistats build in Jenkins - https://phabricator.wikimedia.org/T171599#3489603 (10hashar) [15:58:44] 10Analytics-Kanban, 10User-Elukey: dbstore1002 /srv filling up - https://phabricator.wikimedia.org/T168303#3489608 (10Marostegui) >>! In T168303#3489601, @elukey wrote: > In https://phabricator.wikimedia.org/T170720 we are planning to move one big table (~500GB) to HDFS and drop it from mysql. That would be g... [15:58:56] 10Analytics, 10Contributors-Analysis, 10DBA, 10Chinese-Sites: Data Lake edit data missing for many wikis - https://phabricator.wikimedia.org/T165233#3489609 (10Nuria) @Neil_P._Quinn_WMF The labs snapshot (with newer wikis but not all) is about to start, all things going well it will take 3/4 days for it... [15:59:29] (03PS1) 10Milimetric: Enable newly available wikis for sqooping [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369408 (https://phabricator.wikimedia.org/T165233) [15:59:48] (03CR) 10Milimetric: [V: 032 C: 032] Enable newly available wikis for sqooping [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369408 (https://phabricator.wikimedia.org/T165233) (owner: 10Milimetric) [16:01:25] (03CR) 10Nuria: "Seeing that we are adding new wikis such as jrwiki and frwiki makes me a bit worried about another set of challenges on scoop end." [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369408 (https://phabricator.wikimedia.org/T165233) (owner: 10Milimetric) [16:01:27] (03PS1) 10Milimetric: [WIP] DO NOT MERGE UNTIL THESE WIKIS ARE IMPORTED [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369409 (https://phabricator.wikimedia.org/T165233) [16:02:22] 10Analytics-Kanban, 10Analytics-Wikistats, 10Continuous-Integration-Infrastructure, 10Release-Engineering-Team (Kanban): Fix Wikistats build in Jenkins - https://phabricator.wikimedia.org/T171599#3489630 (10hashar) We now have npm 3.8.3 (the version that came with nodejs 6.0). I have rebuild the job: http... [16:02:49] nuria_: I wouldn't worry about that, because the currently enabled wikis are by far the biggest ones [16:02:54] and we have lots of smaller ones too [16:03:00] milimetric: okeis [16:03:06] * nuria_ hopes for best [16:03:12] none of the newly enabled ones are different in like # of revisions or some other way - but you're right to point it out [16:03:23] I'd bet against problems, but then again I lose a lot of bets :) [16:10:31] 10Analytics, 10Contributors-Analysis, 10DBA, 10Chinese-Sites, 10Patch-For-Review: Data Lake edit data missing for many wikis - https://phabricator.wikimedia.org/T165233#3489668 (10Milimetric) Ok, I just deployed the list for all but the 12 wikis that are still imported. That means the next reconstructio... [16:12:00] 10Analytics-Kanban, 10Patch-For-Review: Add QuickSurvey schemas to EventLogging white-list - https://phabricator.wikimedia.org/T172112#3489669 (10Nuria) [16:20:10] gwicke: from the comments i am not clear ... is the license change ready to go? https://github.com/wikimedia/restbase/pull/848 cc mobrovac [16:21:13] nuria_: still waiting for Zhou's input [16:21:22] gwicke: on github? [16:21:31] no, on the mail thread [16:21:52] and a doc [16:21:57] gwicke: can we have that be a ticket so we have it for future reference when someone from legal asks? [16:22:24] gwicke: there is a phab ticket for this where comments can be posted [16:22:40] gwicke: https://phabricator.wikimedia.org/T170602 [16:22:47] speaking of which, can i be added to the mail thread? [16:22:52] right, it was legal moving it to mail & doc [16:22:53] or better yet, add the services team mail [16:23:10] gwicke: yes, i understand, this is happened before [16:23:16] gwicke: that is why Iam flagging it [16:23:45] and as you probably know I agree with using tasks [16:24:18] gwicke: ya, totally, please do point e-mail thread to doc cause otherwise when questions are asked by someone from legal that is not zhou i do not have place to point them to, this is happen on other license recently [16:24:30] gwicke: would you be so kind as to re-route that to ticket? [16:26:04] gwicke: otherwise it creates quite an overhead for myself (again, this just happened on other license for other data) [16:26:41] just cc'ed you & asked legal to respond on-ticket [16:31:39] gwicke: excellent, thanks [16:34:55] wikimedia/mediawiki-extensions-EventLogging#677 (wmf/1.30.0-wmf.12 - dd54cdd : Antoine Musso): The build has errored. [16:34:55] Change view : https://github.com/wikimedia/mediawiki-extensions-EventLogging/compare/wmf/1.30.0-wmf.12 [16:34:55] Build details : https://travis-ci.org/wikimedia/mediawiki-extensions-EventLogging/builds/259849380 [16:45:52] a-team: https://grafana.wikimedia.org/dashboard/db/analytics-hive [16:48:04] elukey: nice, looking, not sure what heap in hive means though? [16:50:27] elukey: is it one JVM running all queries and heap is tied to that? [16:50:57] nuria_: two jvms, one for the metastore and one for the server (that responds to queries IIUC) [16:51:28] 10Analytics-Kanban, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3489783 (10Nuria) Ping @ZhouZ to leave feedback on ticket [16:58:54] 10Analytics-Kanban: Upgrade Druid to 0.9.2 as a temporary measure - https://phabricator.wikimedia.org/T170590#3489813 (10Ottomata) Hm, building this deb was kind of annoying, because I've already imported 0.10 into the git repo. Here's what I did: ``` git clone ssh://otto@gerrit.wikimedia.org:29418/operations/... [17:00:23] so for example yesterday a big query hit hive (server) and it died because of OOM, so all the clients' requests are stored in the jvm's heap.. More clients means more space consumption, or maybe less queries but heavier.. in this way I hope that we'll be able to track better trends in memory usage [17:00:42] elukey: great [17:03:51] going off now, bye team! [17:03:56] * elukey afk [17:10:20] !log pausing all druid oozie coordinators [17:10:21] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [17:10:27] elukey: fyi, i'm going to try the druid upgrade now [17:10:34] i'm want to make sure this .deb i made works [17:10:42] it'd be weird if i gave it to you and it didnt' [17:24:53] !log beginning druid upgrade to 0.9.2 http://druid.io/docs/0.9.2/operations/rolling-updates.html [17:24:53] Logged the message at https://www.mediawiki.org/wiki/Analytics/Server_Admin_Log [17:59:42] bad news milimetric [17:59:55] some historical loads failed after upgrade [17:59:59] i had to rollback [18:00:03] didn't get very far :/ [18:00:25] looks like it's not going to be so easy [18:00:45] milimetric: i have an idea [18:07:30] 10Analytics-Kanban, 10Patch-For-Review: Upgrade Druid to 0.9.2 as a temporary measure - https://phabricator.wikimedia.org/T170590#3490074 (10Ottomata) Hm, welp, I had to rollback. I only ever restarted the historical node on druid1001. After it finished loading all its historical indexes, a few of them faile... [18:07:52] 10Analytics-Kanban, 10Patch-For-Review: Upgrade Druid to 0.9.2 as a temporary measure - https://phabricator.wikimedia.org/T170590#3490075 (10Ottomata) @elukey, let's hold off on this for now. Upgrading is going to be more delicate than we hoped. [18:29:17] ottomata: no problem, ok, hm... [18:29:32] nuria_: what's up [18:40:54] milimetric: for the upgrade (cc ottomata ) [18:41:12] milimetric: if we see it is not doable in our druid cluster [18:41:31] milimetric: let's just use one of the new druid hosts (thus far outside cluster) [18:41:40] i think won't matter [18:41:42] milimetric: and install druid there ad hoc, newest version [18:41:46] unless we want to use totally different indexes [18:41:51] milimetric: load it with data and test [18:42:06] ottomata: mmm.. what do you mean? [18:42:32] if we want to set it up as a totally different cluster [18:42:33] i guess [18:42:37] ottomata: cause I am suggesting an ad-hoc druid install with version say 9.10 to bea ble to test queries [18:42:41] ottomata: right [18:42:46] ottomata: for testing queries [18:42:56] milimetric, ottomata : it shoudl not matter, right/ [18:42:59] ? [18:43:08] yeah., if new cluster, then it should work [18:43:12] wont' try to load indexes from hdfs [18:43:21] ottomata: actually new ad-hoc host [18:43:40] milimetric: might have its local hdfs install though [18:43:54] ottomata: is the problem the loading of indexes? [18:44:16] ottomata: sorry, i see ticket now, reading [18:45:35] milimetric, ottomata : if you guys are not opposed/ see any problems we could use this one host as a test bed [18:53:07] it works for me, nuria, if ottomata doesn't mind doing it like that [18:53:24] otherwise I can load the data in mysql and test it that way [18:56:56] how will that help though? [18:57:07] if we are going to use druid for this, we need the prod druid to be upgraded, no? [18:57:12] might as well install in labs? [19:10:56] need to test performance, which is probably not possible in labs [19:17:58] 10Analytics-Kanban, 10Patch-For-Review: Oliver Keyes analytics cluster access to check on some old data - https://phabricator.wikimedia.org/T171696#3490516 (10Ottomata) Oliver needed to generate a new key: ``` ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQC2P1KNuHnoFLRFFqwj5cg9qOh5mN3rTz1ei2TTl95FC2toQI2CWGl8b1MK/Oic... [19:22:21] ottomata: no, it is different cause [19:22:29] ottomata: performance will also depend on hardware [19:22:38] cc milimetric [19:22:51] ottomata: so we are not testing just correctness [19:23:11] but you want to test on just one node? sounds like you wanna set upa 3 node secondary druid cluster [19:23:12] :) [19:23:43] ottomata: I do not think we need three to test this data (cc milimetric ) [19:23:49] ottomata: as data is not large in size [19:24:49] but milimetric correct me if i am wrong [19:28:20] nuria: we could try with just one and it'll be obvious if we run into memory problems. But I do expect the performance to be much better with more memory, of course [19:48:30] 10Analytics-Kanban, 10Analytics-Wikistats, 10Continuous-Integration-Infrastructure, 10Release-Engineering-Team (Kanban): Fix Wikistats build in Jenkins - https://phabricator.wikimedia.org/T171599#3490636 (10hashar) [20:35:10] 10Analytics-Kanban, 10Analytics-Wikistats, 10Continuous-Integration-Infrastructure, 10Release-Engineering-Team (Kanban): Fix Wikistats build in Jenkins - https://phabricator.wikimedia.org/T171599#3470314 (10Nuria) It did thank you. Our tests are failing cause it looks like a module is missing (cc @fdans )... [20:41:10] nobody merged anything for the whitelist warning nuria_ [20:41:14] all yours if you want it [20:41:28] milimetric: but how come there are no more warnings? [20:41:39] milimetric: there was only 1 pageview at that time? [20:41:41] maybe it's super low traffic? [20:41:46] yea, possible [20:41:57] you can query the log [20:42:17] I think we have a count there [20:43:13] milimetric: ya, doing that [21:27:32] 10Analytics-Kanban, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491058 (10ZhouZ) I will provide some proposed language we can add. Here's a proposed draft that can be used in the header for the REST API documentation of al... [21:28:42] 10Analytics-Kanban, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491059 (10ZhouZ) I will provide some proposed language we can add. Here's a proposed draft that can be used in the header for the REST API documentation of al... [21:39:24] 10Analytics-Kanban, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491114 (10GWicke) The wording above looks good to me. I think it improves the clarity for all projects. [21:42:23] 10Analytics-Kanban, 10RESTBase-API, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491120 (10mobrovac) The wording looks good to me too, but I have a question. I am probably missing some context, but why does metrics data nee... [22:14:47] 10Analytics-Kanban, 10RESTBase-API, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491216 (10ZhouZ) Hi @mobrovac, without going into the legal weeds, often the license that works best for text content (or software) may not be... [22:17:47] 10Analytics-Kanban, 10RESTBase-API, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491224 (10mobrovac) Oh I see. Thank you for pointing me to the docs, @ZhouZ ! [22:26:46] 10Analytics: Pagecounts-ez not generating - https://phabricator.wikimedia.org/T172032#3491263 (10Erik_Zachte) Working on it. Andrew helped with git problem, also file structure is different on new server. [22:45:00] 10Analytics-Kanban, 10RESTBase-API, 10WMF-Legal, 10Patch-For-Review, 10Services (watching): License for pageview data - https://phabricator.wikimedia.org/T170602#3491298 (10ZhouZ) > The API specification is available under the Apache 2 license. Note it is unclear whether we can make this specific change... [23:22:35] (03PS1) 10Nuria: Adding new wiki to whitelist [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369573 [23:27:42] (03PS2) 10Nuria: Adding new wiki to whitelist [analytics/refinery] - 10https://gerrit.wikimedia.org/r/369573 [23:34:40] seems mailx isn't working on stat1006. were the existing settings (general and per user) imported from stat1003? [23:34:52] tbayer@stat1006:~$ mailx [23:34:52] Cannot open mailbox /var/mail/tbayer: Permission denied [23:45:05] HaeB: this is debian not ubuntu, you can try heirloom-mailx [23:59:23] nuria_: not sure i understand. try out how? which version is currently installed as mailx on stat1006, heirloom or something else? [23:59:45] and how would the switch to debian cause a permissions error?