[10:42:31] qchris: good morning, would you be able to save me with Gerrit please? :-] I need to have the database edited to remove a parent dependency :/ [10:43:14] hashar: Mhmm. Not sure if I can. [10:43:25] But let's hear it. [10:43:37] I am pushing an upstream repository and pushed in Gerrit some upstream commit which created a gerrit change [10:44:01] I have abandoned that change but its commit sha1 is valid and now in the Gerrit DB https://gerrit.wikimedia.org/r/#/c/120761/ [10:44:04] anywa [10:44:18] I then do a merge commit of that upstream branch to our master branch with https://gerrit.wikimedia.org/r/#/c/120763/ [10:44:34] that change has for parent the abandoned commit and thus I can't get it merged [10:44:43] I would need to drop the parent dependency of https://gerrit.wikimedia.org/r/#/c/120763/ [10:45:02] (or get rid of https://gerrit.wikimedia.org/r/#/c/120761/ so Gerrit can't find it and does not insert it as a parent dependency)) [10:45:10] * hashar looks at qchris brain overheating. [10:45:14] :-) [10:45:19] I see the problem. [10:45:27] But I am not sure about the best way to solve it. [10:45:34] Let me think about it for a minute. [10:45:42] another solution is to simply delete the repository integration/composer.git :-] [10:45:58] and I can recreate it. But I don't think we have an easy way to delete a repo [10:46:00] (Gerrit is refreshing several things from git itself, so the issue might reappear upon cache updating ... which we do not want) [10:46:15] deleting a repo is simple. [10:46:32] But you'd loose every change that ever was associated to it. [10:46:37] yeah that is fine :) [10:46:47] Really? [10:46:50] lets takes the lazy/simpler paths [10:46:52] yeah it is ok [10:46:59] Ok. I'll delete it then. [10:47:02] +30 [10:48:18] hashar: integration/composer is gone. [10:48:34] my hero [10:48:47] Should I take care of wiping it from github/gitblit or do you recreate it soonish anyways? [10:48:56] recreating it rightnow [10:48:59] Ok. [10:49:21] you have just saved me a hugeee amount of time. Thank you! [10:49:26] yw. [10:49:33] Glad there was such an easy solution :-) [10:56:40] Gerrit is still p*** me off :-D [11:08:21] milimetric: around ? [12:01:52] matanya, milimetric is on SF this week [12:02:32] thanks nuria, chris told me that [12:02:50] okeis [12:02:57] nuria: if you are here, do you need your access to stat1 ? [12:03:12] i do need access to stat1 yes [12:03:29] are there any changes happening there? [12:03:34] yes [12:03:48] the server is going to move from tampa [12:04:02] can you please move all the stuff you need to stat1002 please ? [12:05:01] ok, will the domain stat1.wikimedia.org change too? [12:05:09] yes, at last [12:05:25] can you please give me a time frame ? [12:06:13] is stat2 [12:06:22] stat1002.wikimedia.org? [12:06:26] yes [12:07:17] same for csalvia ^^ [12:08:07] i cannot ping ping stat1002.wikimedia.org [12:08:39] ..unknown host... [12:09:17] nuria: can you please arrange the move with otto ? [12:10:13] yes, what is what we need to do move whatever we need in stat1 to ? [12:11:23] yes, anything you care of should move from stat1 to stat1002 [12:11:40] you can tell him rt #6789 [12:12:45] Ok, is there an explanation of why i cannot ping stat1002? [12:15:11] from where? [12:16:30] oh, i understand now, nuria it is stat1002.eqiad.wmnet [12:16:31] spain [12:16:37] ah ok [12:19:38] then i guess each of us can move our own personal stuff but still, there are systems on stat1 (like Event Logging) that you guys would be moving .correct? [12:20:22] yes, all in wip [12:20:43] tampa is phasing out, so stuff is in the move phase now [12:21:42] ok, very well, thank you . will move my homedir now [12:23:46] thanks [13:06:14] matanya: Last time I talked about the stat1 move with ottomata, he told me that we don't need to do anything and that [13:06:22] the machine would just get copied over. [13:06:31] Has the plan changed? [13:06:41] Do we need to properly migrate software? [13:07:18] qchris: no, the machine is moving as planned, but we are cleaning access on it [13:07:32] Ah ok :-) [13:07:35] Thanks. [13:07:44] np [13:08:03] Please do not kill my account/data there, as it is needed for geowiki. [14:22:47] (PS15) Milimetric: Add scheduler mode to wikimetrics [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/111623 [15:54:36] http://lists.wikimedia.org/pipermail/glam/2014-March/000581.html "[GLAM] Analytics for Institutional Engagement" [15:55:04] ^^^ someone here probably wants to see that, and/or CC it to your own list. :) [15:57:31] Hi. I was looking for some stats on a beta feature usage across all wikis. [17:10:10] (PS12) Milimetric: Allow reports to be rerun [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/111914 [17:10:25] (PS3) Milimetric: Add utility functions for scheduling [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/119992 [17:13:30] (CR) Milimetric: [WIP] Run recurring reports using the scheduler (1 comment) [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/112165 (owner: Milimetric) [17:50:18] milimetric: hello? [19:49:33] (PS32) Nuria: [WIP] Run recurring reports using the scheduler [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/112165 (owner: Milimetric) [19:49:36] (CR) jenkins-bot: [V: -1] [WIP] Run recurring reports using the scheduler [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/112165 (owner: Milimetric) [20:54:59] qchris, you around? [20:55:06] yes. [20:55:12] yay :) [20:55:17] Looking at gerrit currently :-( [20:55:22] yes yay ;p [20:55:28] So what's up? [20:55:53] So, the mobile requestlogs in Hive; do they have SSL double-counts removed? I appreciate mobile and HTTPS don't go well together usually, but I'm seeing some weird data idiosyncracies and this is a possible hypothesis for where they're coming from [20:57:29] Mhmm. By that you mean that only the request from the SSL terminator to the backend is in Hive, and not the request from the user to the SSL terminator? [20:58:01] I'd assume so, but do not know. [20:58:24] SSL requests come from hosts starting in "ssl" [20:58:49] Those are nginxs, and nginx do not feed into kafka (up to my knowledge) [20:59:03] So I'd assume they are not in Hive. [20:59:37] Do you see requests with hostname starting in "ssl"? [20:59:45] ahh [20:59:54] not that I've seen; looking through the data now :) [21:00:02] TL;DR I got a lot of inter-time periods of 0 seconds. [21:00:06] this struck me as...improbable. [21:00:24] So my hypotheses are (1) my code screwed up (2) my fingerprinting algorithm screwed up (3) double-counting (4) users are stupid. [21:00:39] I do not understand what "inter-time period" mean. What does it mean? [21:01:05] Don't do fingerprinting. It is evil. Do not track users! [21:02:12] yes, yes. [21:02:32] inter-time period; a user makes two 'actual' pageview requests. Subtract the timestamp for the second in seconds from the timestamp of the first. [21:02:36] wait, other way around. Heh ;p. [21:02:49] I get the picture. Thanks. [21:03:08] hrm, looks like it's a fingerprinting problem. Solved! [21:03:16] Oh. Ok :-D [21:03:18] there are still some requests zero seconds apart, but not that many. Cool. [21:03:26] It's a weird world where 'I screwed up' is the good outcome. [21:03:34] Hahaha. [21:03:42] It's learning. [21:03:46] I like learning. [21:03:52] I have to do it a lot. [21:03:58] (= I do screw up often) [22:17:25] milimetric, are you coming to the reddit event tomorrow? [22:17:46] reddit event? :) [22:18:23] will forward [22:19:40] done [22:23:43] http://www.reddit.com/r/Bitcoin/comments/201fa6/hello_from_jimmy_wales_of_wikipedia/ [22:23:48] lzia: related to this ? ^^ [22:25:31] nvm [22:25:34] average, naw, unrelated [22:27:33] lzia: sounds really cool, meeting the reddit folks. I'll keep it in mind and tag along with you guys if I'm not being particularly productive [22:35:03] average, I can't pull up the archives but as Ironholds said, it's unrelated. We had them over and now they have invited. [22:35:15] milimetric, sounds good. [23:11:50] (PS4) Terrrydactyl: Removing cohorts from database. [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/119343 [23:12:13] milimetric: ^