[07:26:06] New patchset: Yuvipanda; "Revert prior 3 commit" [analytics/limn-mobile-data] (master) - https://gerrit.wikimedia.org/r/64885 [15:10:23] ottomata: did you get a chance to run the new country count job? [15:10:35] yup, working on it now [15:10:38] the backfill is running [15:10:40] looking good [15:10:50] looooooooking fine [15:10:59] erosen, i'm changing the metrics.wm.org htpasswd [15:11:05] what is the mailing list email for user metrics? [15:11:38] hmm [15:11:40] one sec [15:11:56] usermetrics@wikimedia.org [15:12:15] is it ok to send the htpasswd to that list? or should I just give it to you and drdee to distribute? [15:12:33] not sure [15:12:43] drdee? [15:12:53] i am setting right next to you [15:12:58] seems like it might be safer to distribute by pm / social network [15:13:02] drdee is sitting right next to me [15:13:08] erosen: can you have a look at 244 output [15:13:18] hmmmmMmmm [15:13:24] and let us know if this is the right format for you to visualize? [15:13:30] erosen, what if I put the pw on stat1 somewhere [15:13:36] I AM SITTiNG RIGHT NEXT TO OTTOMATA! [15:13:38] is there anyone who uses user metrics that doesn't have an account ther? [15:13:45] i think so [15:14:31] maybe maryana? [15:14:35] and swalling? [15:14:42] bwerrrrr [15:14:44] what about fenari? [15:15:01] few people have access to fenari [15:15:06] office wiki is probably better [15:15:16] oh office wiki [15:15:17] ok [15:19:00] ottomata: actually maryana and swalling both have accounts on stat1 [15:19:28] office wiki is fine [15:19:39] erosen: can you look at the output of 244? [15:19:56] yeah, will do [15:22:26] drdee: ottomata: one more thing to consider on this zero job, is that rollups might actually be useful [15:22:34] yup [15:22:41] hmmMMm [15:22:42] i have the logic to load a directory of files [15:22:45] new territory for me, but ok! [15:22:47] but it takes waaaaaay too long [15:22:50] maybe another card? :) [15:23:00] and it takes a really long time to scp to stat1... [15:23:05] minimum viable card [15:23:09] yeah [15:23:29] what kind of rollsup would be useful? [15:23:31] it just means that if we want to fully automate things, we're going to have to throw some bash scripting into things [15:23:38] anything that puts them into a single file [15:23:42] so nothing complicated [15:23:47] even if there are duplicate rows, that's okay [15:34:28] ok erosen, drdee [15:34:30] i just updated the mingle card [15:34:43] zero country and carrier coordinators are running [15:34:50] the new long term regular coordinators start at May 22 [15:35:00] to skip any data weirdness from yesterday/last week [15:35:12] I am running a backfill for country (since it is brand new) from April 1 - May 15 [15:35:23] It'll take a little while to finish that, but its going [15:51:49] drdee: I checked the latest output but there isn't enough data to see anything interesting [15:52:05] it looks like the country job is only there for 1 day: 4/01 [15:52:10] can we start piping the data into the dashboard? [15:52:25] yeah, but the dashboards are aggregated monthly (on amit's request) [15:52:40] and until we get a least a day, we can compute the country fraction graph [15:52:46] its running! [15:52:47] why not show it by day? [15:52:49] **can't** [15:52:56] yeah, actually [15:52:57] erosen [15:52:58] ottomata: why isn't it done!!! ;) [15:53:03] we have such a granular data [15:53:08] do you want me to do this on a day level instead of by hour? [15:53:10] we can [15:53:18] the backfill will then go faster [15:53:31] there is no reason to do hourly that I know of [15:53:44] amit has never asked for it / mentioned it [15:53:52] hm ooooookkkk [15:53:57] let's do it daily [15:54:09] sounds good to me [16:03:59] erosen, if I do this daily, i'm going to completely redo carrier then as well [16:04:12] that's fine with me [16:15:56] ok, looks like daily is working [16:16:02] its backfilling both now [17:59:38] rounce123: I can't make it to the analytics review today, in a meeting with Kirsten for UMAPI/EventLogging docs [18:00:01] ping me if you need me [18:00:08] No problem most of the team is out so we are not having it today [18:00:20] ok cool [18:00:46] Will re-group next week but you can see the work that is planned on the WIP board. Please reach out if you have any questions [18:09:32] Curious pattern on this graph: http://j.mp/14RQ24V [18:09:44] Would anyone know why there's a bump down at the top of every hour? [18:18:15] niemeyer: I think ottomata would be the best person to ask about this. He is in transit at the moment, but i suspect he would appreciate an e-mail. [18:19:44] erosen: Thanks, will send it. What's his email? [18:19:47] or her [18:30:14] niemeyer: sorry for the lapse. his e-mail is otto@wikimedia.org [18:30:18] thanks again for pointing this out [18:30:25] erosen: Cheers [19:13:10] Change merged: JGonera; [analytics/limn-mobile-data] (master) - https://gerrit.wikimedia.org/r/64885 [20:36:21] I'm curious about why shutting down just a few of the app servers affected the incoming network traffic so dramatically: http://j.mp/14yRfyw [22:17:27] Change merged: JGonera; [analytics/limn-mobile-data] (master) - https://gerrit.wikimedia.org/r/62664