[08:13:01] <grrrit-wm>	 (CR) Nuria: "Retested on dev and things are working with cohort, report and symlink creation" [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/143040 (https://bugzilla.wikimedia.org/66087) (owner: Milimetric)
[08:13:13] <grrrit-wm>	 (CR) Nuria: [C: 2] Add pretty symlink for WikimetricsBot [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/143040 (https://bugzilla.wikimedia.org/66087) (owner: Milimetric)
[08:13:24] <grrrit-wm>	 (Merged) jenkins-bot: Add pretty symlink for WikimetricsBot [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/143040 (https://bugzilla.wikimedia.org/66087) (owner: Milimetric)
[08:16:05] <grrrit-wm>	 (CR) Nuria: "I actually think we should not merge these changes until we have done some load testing and know the report rate creation we can sustain." [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/142007 (https://bugzilla.wikimedia.org/66841) (owner: Milimetric)
[10:19:46] <nuria>	 all right qchris: newly register metrics runs  on dev work well: https://metrics-dev.wmflabs.org/static/public/datafiles/NewlyRegistered/
[10:20:19] <qchris>	 k
[10:20:42] <qchris>	 "502 Bad Gateway"
[10:20:56] <qchris>	 nuria: ^ is what the above url gives for me.
[10:21:16] <nuria>	 really?
[10:21:20] <qchris>	 yup.
[10:21:21] <nuria>	 works like a charm
[10:21:27] <nuria>	 for me
[10:21:36] <nuria>	 seems labs avilability issue
[10:21:43] <nuria>	 "labs availability
[10:21:58] <qchris>	 Wait ... it works in wget, but not in firefox.
[10:22:12] <nuria>	 juas
[10:22:40] <qchris>	 Now it works in firefox. ... Maybe a nginx hiccup?
[10:23:18] <nuria>	 ya
[10:23:37] <nuria>	 i suspect the load we can sustain there is nothing
[11:23:20] <milimetric>	 morning qchris
[11:23:29] <milimetric>	 so, thanks so much for fixing labs
[11:23:30] <qchris>	 milimetric: Hi
[11:23:38] <qchris>	 yw
[11:23:50] <milimetric>	 secondly, I was going to test my oozie flow and coordinators
[11:24:01] <milimetric>	 Andrew said CDH5 should be up and running
[11:24:08] <milimetric>	 and I know you were going to look at labs
[11:24:21] <qchris>	 I read him saying that you can use his "e" cluster.
[11:24:30] <milimetric>	 yea
[11:24:48] <milimetric>	 so of course I'm not sure what that means and I was thinking you did :)
[11:25:31] <qchris>	 So you if you wanted to test oozie, just log in into one of the instances, like "hadoop-e-worker0"
[11:25:44] <qchris>	 and you can use that if it were a machine of our production cluster.
[11:25:56] <qchris>	 Hive & Co should be available.
[11:26:02] <milimetric>	 k, trying
[11:26:10] <qchris>	 IIRC he suggested an instance to use ... let me read the backlog.
[11:29:07] <qchris>	 Cannot find it ottomata's suggestion ... but "hadoop-e-worker0" has client roles for oozie and hive. So that instance should work.
[11:41:57] <milimetric>	 argh, qchris: can I ask you questions when I get stuck?  or would that be annoying at the moment?
[11:43:27] <qchris>	 milimetric: Sure. Ask questions. Not sure though if I can help. But first ... lunch.
[11:43:31] <qchris>	 Sorry.
[11:51:27] <milimetric>	 :( sigh, this is super sad, but I give up, I can't find it
[11:51:33] <milimetric>	 I need the create table for webrequest
[11:51:39] <milimetric>	 but the cluster's down so I can't get it there
[11:51:54] <milimetric>	 and I searched all over puppet and couldn't find it there
[11:57:06] <YuviPanda>	 milimetric: if you're looking for cdh5 on puppet, you should look into the cdh5 submodule, it's not in the main repo
[11:57:25] <milimetric>	 oh, I know, i looked through:
[11:57:37] <milimetric>	 operations/puppet/cdh (new version of ../../cdh4)
[11:57:46] <milimetric>	 operations/puppet/varnishkafka
[11:57:49] <milimetric>	 operations/puppet/kafka
[11:58:10] <milimetric>	 but yeah, those were all long shots as I wouldn't expect this to be in there
[11:59:10] <milimetric>	 i would just always grab it off the cluster so now that it's down I'm all confused
[12:00:05] * milimetric doesn't understand ops work, he thinks it's all magic and sorcery
[12:11:50] <milimetric>	 closest I got so far:
[12:11:51] <milimetric>	     # We are only using hive-partition to add partitions to the pagecounts table.
[12:11:51] <milimetric>	     # The webrequest table is using Oozie.
[12:11:51] <milimetric>	     $tables      = 'pagecounts'
[12:42:34] <qchris>	 milimetric: Are you sure the webrequest table is created through puppet? I thought this was done by hand ...
[12:42:43] <milimetric>	 qchris: no clue :)
[12:43:00] <milimetric>	 I was hoping it was done in some way I could find, so I can get the schema
[12:43:03] <qchris>	 So if you need the table in a labs cluste, just create it by hand.
[12:43:15] <milimetric>	 right, but I don't know the schema
[12:43:24] <qchris>	 For the oozie stuff, you only need some columns.
[12:43:42] <milimetric>	 mmm, you're saying just ignore the other stuff.... mmm, ok
[12:44:06] <qchris>	 I'd do it that way ... but then again ... maybe I am just sloppy.
[12:45:49] <qchris>	 milimetric: I know that https://wikitech.wikimedia.org/wiki/Analytics/Kraken/Hive/Schemas
[12:45:58] <qchris>	 is outdated, but it might serve as start.
[12:46:27] <qchris>	 (outdated, because at least the webrequest_source column is missing)
[12:47:13] <milimetric>	 no, I'm dumb, I can just create the columns I need to test and not be nitpicky about stuff that doesn't matter
[12:48:35] <qchris>	 http://dpaste.com/3YKS2ZP
[12:48:50] <qchris>	 ^ is the lasted version I have locally. It is from 2014-05-29
[12:49:14] <qchris>	 I assume that it is still accurate (except the location setting)
[12:49:17] <qchris>	 milimetric: ^
[12:49:41] <milimetric>	 thanks :)
[12:50:11] <qchris>	 yw
[13:22:35] <milimetric>	 qchris: I get permission denied trying to insert to a table in hadoop
[13:22:52] <milimetric>	 also - hive syntax for inserting ad-hoc is insane :)
[13:23:11] <qchris>	 As which user? "milimetric"?
[13:23:16] <milimetric>	 yes
[13:23:17] <qchris>	 Or "hive" "hdfs".
[13:23:19] <qchris>	 Ok.
[13:23:26] <milimetric>	 should i become another user?
[13:23:35] <qchris>	 Not sure ... might be.
[13:24:10] <qchris>	 Let me check the labs instance ...
[13:26:52] <milimetric>	 ok qchris, hdfs user works
[13:26:55] <milimetric>	 (sudo su hdfs)
[13:26:58] <qchris>	 ok.
[13:27:00] <milimetric>	 so last problem
[13:27:15] <milimetric>	 inserting ad-hoc in hive requires at least one row to exist in at least one table somewhere
[13:27:27] <milimetric>	 is there an easier way other than creating an external table?
[13:27:31] <qchris>	 yup. That sucks big time.
[13:27:42] <qchris>	 I typcially create my own database.
[13:27:49] <qchris>	 Create a simple table on that database.
[13:27:51] <milimetric>	 right, i created wmf
[13:27:54] <milimetric>	 then webrequest in there
[13:27:55] <qchris>	 With one column.
[13:28:02] <qchris>	 Add a single row to that table.
[13:28:03] <milimetric>	 oh ok, and you make that table external
[13:28:17] <qchris>	 And then, you can use that table to create into another table.
[13:28:25] <milimetric>	 how do you "add a single row"
[13:29:04] <qchris>	 insert either from a local file, or select from a different table (hence the single row-ed separate table I typically create)
[13:29:22] <qchris>	 (you can also insert from a file on hdfs, but that's even stranger)
[13:30:36] <qchris>	 Ha ... Stack Overflow even suggests the same thing: http://stackoverflow.com/questions/17425492/hive-insert-query-like-sql
[13:31:17] <milimetric>	 yea, I'm not sure how to do any of those things (insert from file or insert from file on hdfs)
[13:31:36] <milimetric>	 select from a different table isn't possible because there is no other table on the cluster
[13:32:19] <qchris>	 Ok. I'll bootstrap the table.
[13:33:24] <milimetric>	 oh no, i see the SO answer has example
[13:33:36] <milimetric>	 i got it, sorry i didn't see - the cool answer is buried down
[13:33:43] <milimetric>	 qchris: ^
[13:33:52] <qchris>	 Ok.
[13:39:22] <qchris>	 milimetric: Going through my test scripts I found a line that matches the schema I dpasted before.
[13:39:31] <milimetric>	 cool
[13:39:36] <qchris>	 You should be able to "load data local" from /tmp/logline.txt on
[13:39:43] <qchris>	 hadoop-e-worker0
[13:39:52] <qchris>	 (Just copied the line there)
[13:40:14] <qchris>	 (It is cleaned from PII, so just go wild with it)
[13:56:00] <milimetric>	 so weird
[13:56:14] <milimetric>	 i put demo.txt in /hadoop-e/user/milimetric/ and it won't work from there
[13:56:15] <milimetric>	 trying yours
[13:56:28] <milimetric>	 FAILED: SemanticException Line 1:17 Invalid path ''demo.txt'': No files matching path hdfs://hadoop-e/user/hdfs/demo.txt
[13:56:38] <milimetric>	 hdfs@hadoop-e-worker0:/home/milimetric$ hdfs dfs -ls /hadoop-e/user/milimetric
[13:56:38] <milimetric>	 Found 1 items
[13:56:38] <milimetric>	 -rw-r--r--   3 hdfs hadoop          2 2014-07-04 13:38 /hadoop-e/user/milimetric/demo.txt
[13:57:25] <milimetric>	 hm, yours doesn't work either, I must be don't something wrong qchris:
[13:57:26] <milimetric>	 hive (wmf)> load data inpath '/tmp/logline.txt' into table webrequest;
[13:57:26] <milimetric>	 FAILED: SemanticException Line 1:17 Invalid path ''/tmp/logline.txt'': No files matching path hdfs://hadoop-e/tmp/logline.txt
[13:57:59] <qchris>	 You're missing the "local"
[13:58:09] <qchris>	 '<qchris> You should be able to "load data local" from /tmp/logline.txt on'
[13:58:21] <qchris>	 Without it, hive tries to load it from hdfs.
[13:58:50] <qchris>	 I'll jump into standup meeting, and afterwards, we get the table bootstrapped.
[14:31:26] <milimetric>	 qchris: what's localhost stand for in:
[14:31:27] <milimetric>	 oozie job -oozie http://localhost:8080/oozie -config job.properties -run
[14:31:44] <qchris>	 typically 127.0.0.1.
[14:31:57] <qchris>	 Ubuntu has something slightly different in /etc/hosts IIRC.
[14:32:00] <milimetric>	 i mean, in our case, is oozie on local:8080?
[14:32:53] <qchris>	 Ahm. Not sure.
[14:33:25] <qchris>	 Telnet fails to connect on localhost:8080
[14:33:32] <milimetric>	 yeah, apparently nothing's serving there
[14:33:41] <milimetric>	 hm, so where do we think oozie is here... hmmm
[14:34:22] <qchris>	 hadoop-e-master0  is oozie server.
[14:34:25] <qchris>	 milimetric: ^
[14:34:34] <milimetric>	 ah!
[14:35:05] <qchris>	 But ... 8080 for oozie does not sound plausible.
[14:35:45] <qchris>	 Default for "-oozie" seems to work just fine.
[14:35:51] <qchris>	 So I'd not specify it.
[14:35:57] <milimetric>	 oh really?
[14:35:58] <milimetric>	 dammit
[14:36:15] <qchris>	 I just ran"oozie jobs" and itgave me nice output.
[14:36:22] <milimetric>	 sorry then
[14:36:26] <qchris>	 no worries.
[14:37:47] <milimetric>	 hm, no, -oozie command needs it:
[14:37:48] <milimetric>	 milimetric@hadoop-e-master0:~$ oozie job -oozie -config refinery/oozie/webrequest/sequence_stats/workflow.properties -run
[14:37:48] <milimetric>	 Invalid sub-command: Missing argument for option: oozie
[14:37:48] <milimetric>	 use 'help [sub-command]' for help details
[14:37:48] <milimetric>	 milimetric@hadoop-e-master0:~$ oozie job -oozie http://localhost:8080/oozie -config refinery/oozie/webrequest/sequence_stats/workflow.properties -run
[14:37:48] <milimetric>	 Error: IO_ERROR : java.net.ConnectException: Connection refused
[14:38:58] <qchris>	 Sorry, I was too value :-(  You gave "-oozie" but no value for it. What about not even passing "-oozie"
[14:39:06] <qchris>	 The oozie will use the default.
[14:39:27] <qchris>	 So something like: oozie job  -config refinery/oozie/webrequest/sequence_stats/workflow.properties -run
[14:42:04] <milimetric>	 yeah, no i'm an idiot
[14:42:05] <milimetric>	 it's official
[14:42:38] <qchris>	 No. The hadoop & ecosystem CLI tools suck. Really.
[14:42:57] <qchris>	 Every tool is a bit different.
[14:43:05] <qchris>	 :-(
[14:47:24] <milimetric>	 hm, well maybe we can help improve them at some point in the distant future, as we've certainly felt the pain
[14:47:39] <milimetric>	 so - now the job didn't error
[14:47:42] <milimetric>	 but it's just sitting there hanging
[14:47:46] <qchris>	 Sounds great \o/
[14:47:59] <qchris>	 Oh wait ... hanging :-D
[14:48:02] <milimetric>	 that can't be right, i mean there's 2 records in  the table :)
[14:48:24] <qchris>	 For the period of time we're interested in?
[14:48:35] <milimetric>	 no, like i put only 2 records in the whole table
[14:48:39] <milimetric>	 :D
[14:48:59] <milimetric>	 so something's not working if the job's just taking like 10 minutes to start
[14:49:09] <milimetric>	 it doesn't show up in `oozie jobs` yet
[14:49:33] <qchris>	 That's what we love oozie for.
[14:49:50] <milimetric>	 yep.  ok, it's clear to me I'm not getting any progress on this.  I'm going to switch to wikimetrics
[14:49:58] <qchris>	 Haha :-)
[14:50:00] <qchris>	 k.
[14:50:07] <milimetric>	 thanks for all the great help though
[14:50:17] <qchris>	 Hey ... you could help me in return ...
[14:50:25] <qchris>	 Can you see instances on:
[14:50:38] <qchris>	 https://wikitech.wikimedia.org/wiki/Special:NovaInstance
[14:50:54] <qchris>	 All projects look empty for me since some hours.
[14:52:26] <qchris>	 I guess I'll just ask in the labs channel
[15:08:16] <milimetric>	 qchris: yes, saw the same thing
[15:08:32] <qchris>	 Ok. Thanks.
[15:08:41] <milimetric>	 ah! ping me I got all excited about being productive again :)
[15:09:00] <qchris>	 Oh .. wait "saw� ... does it mean it's back to normal now for you?
[15:09:32] <milimetric>	 yes qchris, sorry
[15:09:40] <milimetric>	 i was t rying to read in labs if you already got it figured out
[15:09:46] <milimetric>	 i'll answer there for others' benefit
[15:10:11] <qchris>	 ok. please do. Thanks.
[15:18:07] <milimetric>	 oh, gerrit question:
[15:18:27] <milimetric>	 I referenced a commit the other day from analytics/wikimetrics in a ops/puppet/wikimetrics change
[15:18:38] <milimetric>	 and you reverted it I think and connected them some other way?
[15:18:43] <milimetric>	 or am I misremembering
[15:19:01] <milimetric>	 I'm about to do the same thing, and just wanted to ask first
[15:19:03] <qchris>	 Sorry ... I do not recall. Do you have urls for me?
[15:19:12] <milimetric>	 it's not important, but if I have:
[15:19:21] <milimetric>	 wikimetrics change <-- should reference --> puppet change
[15:19:29] <milimetric>	 I'm about to just put that in the commit message of each one
[15:19:34] <milimetric>	 is that bad?
[15:19:53] <qchris>	 Not necessarily.
[15:20:23] <qchris>	 Using sha1 of other commits in a commit message is fine if the things are merged already.
[15:20:33] <qchris>	 Then git tools can find them automatically.
[15:20:43] <milimetric>	 I was going to just use a link to the gerrit repo
[15:20:55] <milimetric>	 *gerrit change
[15:21:01] <qchris>	 But for things that are not yet merged, I'd rather use change numbers, because a rebase changes tho SHA1 but does not change the change number.
[15:21:14] <milimetric>	 ok, so change number but not link to change?
[15:21:25] <milimetric>	 like not https://gerrit.wikimedia.org/r/#/c/142007/ but Change: 142007
[15:21:26] <milimetric>	 ?
[15:21:30] <nuria>	 holaaa
[15:21:33] <milimetric>	 hi nuria
[15:21:33] <qchris>	 No, use the link.
[15:21:40] <milimetric>	 oh ok, will do, thanks!
[15:21:50] <qchris>	 Hi nuria
[15:22:04] <nuria>	 hello, i though milimetric was on va-ca-ti-on
[15:22:21] <nuria>	 on holiday, rather
[15:23:00] <grrrit-wm>	 (PS4) Milimetric: Remove limit on recurrent, add throttling [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/142007 (https://bugzilla.wikimedia.org/66841)
[15:23:10] <milimetric>	 :)
[15:23:54] <milimetric>	 qu'est-ce que c'est vacation?
[15:24:20] <milimetric>	 no, just kidding, I wanted to at least try to work on the oozie stuff
[15:24:23] <nuria>	 juas
[15:24:25] <milimetric>	 I failed miserably but I at least tried
[15:24:41] <milimetric>	 but! nuria: perfect timing
[15:24:48] <nuria>	 i can take the oozie stuff or do we need andrew for that
[15:24:49] <milimetric>	 you said yesterday that "365" was kind of arbitrary
[15:24:50] <milimetric>	 I agree: https://gerrit.wikimedia.org/r/#/c/142007/
[15:25:09] <milimetric>	 the last patch adds a configuration setting, and I added it to puppet just now
[15:25:43] <nuria>	 k let me look at it
[15:25:49] <milimetric>	 i think we need andrew nuria, oozie refuses to accept my job
[15:50:41] <nuria>	 mmm.. i get a bunch of integrity errors when trying to create that many reports (100) on vagrant, let me see if i can see what is going on
[15:52:57] <nuria>	 I tested batches of 60 ealier on today w/o issues though
[15:54:26] <milimetric>	 weird
[15:54:28] <milimetric>	 what's the error?
[15:54:49] <nuria>	 Problem creating child report: (IntegrityError) (1062, "Duplicate entry '23-2014-06-02 00:00:00' for key 'uix_report'") 'INSERT INTO report (created, user_id, queue_result_key, result_key, status, name, show_in_ui, parameters, public, recurrent, recurrent_parent_id) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)' (datetime.datetime(2014, 6, 2, 0, 0), 1L, None, None, 'PENDING', None, 0,
[15:54:49] <nuria>	  '{\n    "cohort": {\n        "id": 8, \n        "name": "wiki", \n        "size": 0\n    }, \n    "metric": {\n        "name": "NewlyRegistered", \n        "end_date": "2014-06-02 00:00:00", \n        "start_date": "2014-06-01 00:00:00"\n    }, \n    "recurrent": false, \n    "public": true, \n    "name": "wiki - NewlyRegistered"\n}', 1, 0, 23L)
[16:05:53] <milimetric>	 yeah, that index is on recurrent_parent_id and created, so it sounds like a real problem.  But I don't know how it got in that state so it's hard to tell.  Maybe the new logic doesn't properly select the old already-finished runs
[16:06:03] <milimetric>	 what happens when you:
[16:06:26] <milimetric>	 select (* minus parameters) from report where recurrent_parent_id = 23;
[16:16:49] <nuria>	 milimetric, all reports look to be there ...
[16:17:44] <milimetric>	 right, so it shouldn't try to run any more, and it sounds like your instance IS trying to run more
[16:17:55] <milimetric>	 I'm finishing up a quick patch and I'll take a look
[16:20:08] <nuria>	 also "Problem creating child report: QueuePool limit of siz
[16:20:08] <nuria>	 e 5 overflow 10 reached, connection timed out, timeout"
[16:20:22] <nuria>	 this hhapens when trying to create ~90 reports on vagrant
[16:20:26] <grrrit-wm>	 (PS1) Milimetric: Fix test that adds existing tag [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/144159 (https://bugzilla.wikimedia.org/66671)
[16:20:31] <wikibugs>	 Analytics / Wikimetrics: test_add_new_tag_utf8 fails - https://bugzilla.wikimedia.org/67166#c2 (Dan Andreescu) NEW>RESO/INV I couldn't replicate this, and I may have been a bit confused when it was first reported, because of another bug: https://bugzilla.wikimedia.org/show_bug.cgi?id=66671.  Eith...
[16:20:40] <nuria>	 up to that all works welll
[16:21:30] <milimetric>	 huh, ok, so nuria, here's the steps I'll try to reproduce, let me know if I'm missing anything:
[16:21:38] <milimetric>	 rebuild vagrant
[16:21:44] <milimetric>	 (hold on let me start that)
[16:22:04] <nuria>	 you know, the thing is that dev is a better testing ground
[16:22:41] <nuria>	 cause in vagrant limits are going to be different
[16:23:06] <nuria>	 so i would test intervals there ( i alredy created batches of 60 in dev w/o issues)
[16:23:39] <nuria>	 i changed the queue.yaml in dev to execute on 'debu' mode ( and thus the scheduler runs more often)
[16:23:47] <nuria>	 sorry, debug mode
[16:24:06] <nuria>	 we can try the logic there
[16:26:05] <milimetric>	 ok, I just got done rebuilding vagrant
[16:26:13] <milimetric>	 seriously, I have no idea how you guys like this :)
[16:26:23] <milimetric>	 it triplicates the work I have to do every time I touch it
[16:26:26] <milimetric>	 anyway
[16:26:40] <milimetric>	 well, nuria, why don't we test in staging?
[16:26:57] <nuria>	 sure, it's the same
[16:27:03] <milimetric>	 ah, vagrant
[16:27:06] <milimetric>	 's broken anyway: Could not parse options: invalid option: --hiera_config=/tmp/vagrant-puppet-1/hiera.yaml
[16:27:08] <nuria>	 right? dev and staging are identical virt machines
[16:27:22] <milimetric>	 well, but you already did some other stuff in dev
[16:27:32] <milimetric>	 i guess we could just wipe the dbs and reset everything
[16:27:36] <milimetric>	 oh, do you wanna hangout and do it?
[16:27:40] <nuria>	 in this case vagrant will be missleading
[16:27:42] <nuria>	 sure
[16:27:51] <nuria>	 give me 2 mins
[16:33:35] <nuria>	 ok, reday
[16:33:39] <nuria>	 *ready
[16:57:47] <wikibugs>	 Analytics / Wikimetrics: Wikimetrics needs to read config files /etc/wikimetrics/ rather that files on local repository by default when setting up the config - https://bugzilla.wikimedia.org/67542 (nuria) NEW p:Unprio s:normal a:None Wikimetrics needs to read yaml config files in /etc/wikim...
[17:42:43] <grrrit-wm>	 (PS5) Milimetric: Remove limit on recurrent, add throttling [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/142007 (https://bugzilla.wikimedia.org/66841)
[17:57:04] <grrrit-wm>	 (PS6) Milimetric: Remove limit on recurrent, add throttling [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/142007 (https://bugzilla.wikimedia.org/66841)
[18:04:01] <wikibugs>	 Analytics / EventLogging: Fix per schema counts in graphite - https://bugzilla.wikimedia.org/67073 (Dan Andreescu) PATC>RESO/FIX
[18:04:16] <wikibugs>	 Analytics / Wikimetrics: Story: CronUser runs EEVS metrics on all projects daily - https://bugzilla.wikimedia.org/65946 (Dan Andreescu) PATC>RESO/FIX
[18:04:30] <wikibugs>	 Analytics / Wikimetrics: Path to recurrent report (from dashboard) should have better semantics - https://bugzilla.wikimedia.org/66087 (Dan Andreescu) PATC>RESO/FIX
[18:08:16] <wikibugs>	 Analytics / Wikimetrics: Story: CronUser backfills NewlyRegistered User for recurring reports - https://bugzilla.wikimedia.org/66841#c3 (Dan Andreescu) PATC>RESO/FIX tested in staging, most everything looking good, but we did find a performance issue we're going to follow up on with: https://bugz...
[18:08:18] <wikibugs>	 Analytics / Wikimetrics: Performance of Recurrent Reports - https://bugzilla.wikimedia.org/67543 (Dan Andreescu) NEW p:Unprio s:normal a:None While testing how many instances of recurring reports we could spawn as part of the daily run, we discovered a lack of understanding around wikimetri...
[18:08:48] <milimetric>	 hey look at that we're on-track: http://sb.wmflabs.org/t/analytics-developers/2014-06-26/
[19:14:03] <nuria>	 magically...
[19:34:38] <Guest91629>	 [travis-ci] develop/f709bb6 (#176 by nuria): The build passed. http://travis-ci.org/wikimedia/limn/builds/29165039