[08:34:35] <ori>	 nuria: i deployed it, it was trivial
[08:34:43] <nuria>	 ah ya, alredy?
[08:34:46] <nuria>	 *already?
[08:34:58] <nuria>	 ok, will look at graphite within the day today
[08:35:14] <ori>	 nuria: http://ganglia.wikimedia.org/latest/graph.php?r=day&z=xlarge&c=Miscellaneous+eqiad&h=vanadium.eqiad.wmnet&jr=&js=&v=6.2&m=cpu_user&vl=%25&ti=CPU+User
[08:35:18] <ori>	 can you guess when? ;)
[08:35:24] <nuria>	 whattttt???
[08:35:49] <nuria>	 ok-> you also get 'sinny star of the week award'
[08:36:04] <ori>	 sinny?
[08:37:03] <nuria>	 sorry "Shining Star"
[08:37:06] <ori>	 heh
[08:37:26] <ori>	 validating schema is more expensive than validating events, because the meta-schema (that schema have to be validated against) is complicated
[08:37:39] <ori>	 we were already doing it prior to caching the schema locally
[08:38:12] <ori>	 https://github.com/wikimedia/mediawiki-extensions-EventLogging/blob/master/server/eventlogging/schema.py#L60
[08:39:27] <nuria>	 i love it when fixes make sense
[08:39:58] <ori>	 fixed perf problems by deleting a single line!
[08:40:04] <ori>	 one that i stupidly put in, but nevermind that
[08:41:23] <ori>	 we should still spread validation over multiple workers but we bought ourselves a lot of time
[08:44:11] <nuria>	 puff .. if all perf fixes were like this one ...
[08:45:43] <nuria>	 what I still do not get is this graph: http://ganglia.wikimedia.org/latest/graph_all_periods.php?h=vanadium.eqiad.wmnet&m=cpu_report&r=hour&s=descending&hc=4&mc=2&st=1402044018&g=cpu_report&z=large&c=Miscellaneous%20eqiad
[08:46:11] <nuria>	 max is 25% because there are 4 cores?
[08:47:07] <nuria>	 but ....how is load distributed across cores w/o threading there?
[08:50:50] <ori>	 it isn't
[08:50:55] <ori>	 the os runs different processes on different cores
[08:51:04] <ori>	 but a single python process can't utilize more than one
[09:27:47] <nuria>	 tgr: I believe uuids are not integerts but rather they are of this form: "uuid": "1a9d398809fc5db1aac3ba9d1ec1d7ed"
[10:12:44] <travis-ci>	 [travis-ci] wikimedia/mediawiki-extensions-EventLogging#211 (master - 3d8d20f : Ori Livneh): The build was broken.
[10:12:44] <travis-ci>	 [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-EventLogging/compare/f9ce0378a778...3d8d20fd37a9
[10:12:44] <travis-ci>	 [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-EventLogging/builds/26927603
[16:08:05] <ottomata>	 ah poo christian
[16:08:08] <ottomata>	 qchris_away:
[16:08:14] <ottomata>	 i accidentally pushed this directly
[16:08:14] <ottomata>	 https://github.com/wikimedia/analytics-camus/commit/115e40a973585cc8cdfc2fe8594b9e22ae7d238a
[16:08:25] <ottomata>	 meant to push it for review, but was having trouble with merge history
[16:08:42] <ottomata>	 good news though, it compiles!
[16:27:10] <ottomata>	 Ironholds: hiyaaaa
[16:28:24] <qchris_away>	 ottomata: Do we do code-review around camus? Anyways ... fine by me.
[16:28:25] <qchris_away>	 Looks good.
[16:28:28] <ottomata>	 we do for wmf things
[16:28:34] <qchris_away>	 Oh ok.
[16:28:39] <ottomata>	 the commits that are meant to stay in just the wmf branch
[16:28:47] <ottomata>	 the ones that get upstreamed are code reviewed via pull requests :/
[16:48:45] * Ironholds sighs
[16:48:45] <ottomata>	 camus runs and compiles in cdh5, woooo!
[16:48:52] <ottomata>	 Ironholds: hiya!
[16:48:52] <Ironholds>	 a day and a half on a reduce task :(
[16:48:54] <Ironholds>	 ottomata, yay!
[16:48:57] <ottomata>	 your thingee is still running?
[16:48:59] * Ironholds dances
[16:49:04] <Ironholds>	 it is, but more WOO CDH5!
[16:49:09] <ottomata>	 haha
[16:49:21] <Ironholds>	 we can have UDFs in the query, and TABLESAMPLE, and a pony, and and and...
[16:50:08] <ottomata>	 Ironholds: can I delete 2014/04 webrequest_mobile data?
[16:50:31] <ottomata>	 ..brb
[16:51:59] <Ironholds>	 ottomata, totally.
[19:32:30] <Ironholds>	 2014-06-06 16:49:16,754 Stage-1 map = 100%,  reduce = 40%, Cumulative CPU 4052190.35 sec
[19:32:30] <Ironholds>	 2014-06-06 16:49:18,471 Stage-1 map = 100%,  reduce = 0%,
[19:32:33] <Ironholds>	 what in god's name...
[19:34:41] <ottomata>	 ha
[19:40:41] <qchris>	 Ironholds: I noticed that also in the log you uploaded to the bug.
[19:40:59] <Ironholds>	 yeah. Is this a thing, or...?
[19:41:11] <Ironholds>	 this is a different query that has previously, happily completed.
[19:41:45] <qchris>	 Mhmm. I have no clue. I never saw this happen for the queries I run.
[19:42:02] <qchris>	 (But they were probably smaller than the ones you run)
[19:53:44] <Ironholds>	 hurm
[19:53:56] <Ironholds>	 there seems to be a problem with the geolocation element of the NavigationTiming schemas
[19:54:01] <ottomata>	 Ironholds: sorry, i just noticed that you updated that bug with your app id 2 days ago
[19:54:07] <Ironholds>	 ottomata, np!
[19:54:14] <ottomata>	 i think i do not get bugs emails very well, i must have an email filter problem
[19:54:29] <ottomata>	 if there's something you want me to look into sooner rather than later, and you've updated a bug and nohting is moving, do feel free to ping me!
[19:54:36] <Ironholds>	 sure!
[19:54:40] <Ironholds>	 I had enough work to keep me busy :)
[20:52:42] <grrrit-wm>	 (PS1) Nuria: [WIP] Adding test for empty user results on metric [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/138031
[20:52:54] <grrrit-wm>	 (CR) jenkins-bot: [V: -1] [WIP] Adding test for empty user results on metric [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/138031 (owner: Nuria)
[23:38:01] <grrrit-wm>	 (PS1) Nuria: [WIP] 1st run of recurrent report is scheduled right away [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/138109 (https://bugzilla.wikimedia.org/66280)