[01:51:52] 6Labs, 3Labs-Sprint-107, 3Labs-Sprint-108: Evaluate kubernetes for use on Tool Labs - https://phabricator.wikimedia.org/T107993#1513457 (10yuvipanda) Ok, so there is now an experimental 'k8s' puppet module, which is applied on all hosts in k8s-eval project. It has docker running with flannel overlay network... [03:58:44] 6Labs, 3Labs-Sprint-107, 3Labs-Sprint-108: Evaluate kubernetes for use on Tool Labs - https://phabricator.wikimedia.org/T107993#1513561 (10yuvipanda) Note that you might have to use `brctl delbr docker0` and restart docker to get it to accept flannel - this is because of missing dependency chains that we oug... [04:21:03] YuviPanda: ha I see docker stuff happening! [04:21:38] Negative24: indeed :) and it is me doing it now [04:21:47] YuviPanda: but I want in [04:24:20] YuviPanda: could you review https://phabricator.wikimedia.org/T106969 again? [04:24:21] T106969: Labs Docker Project [04:24:32] whoops though I disabled that [04:24:39] Bot24: role phablookup stop [04:24:39] Stopping phablookup... [04:25:05] Bot24: role phablookup status [04:25:07] phablookup is off [04:29:24] YuviPanda: docker might be an interesting platform on tools /crazy idea [04:43:52] Negative24: the long term evil plan® is to make something a lot like heroku to replace the current grid engine system [04:44:38] light weight scalable services that are magically managed by pushing to special git repos [04:44:38] bd808: yep that's similar to what I was thinking [04:46:59] Negative24: you are interested in docker for MW? [04:47:08] * bd808 just read linked ticket [04:47:17] bd808: yes and other stuff [04:47:57] Docker has often been mentioned as a magic solution of packaging MW plus services [04:48:04] I'm not totally convinced yet [04:48:20] running a docker container with a full init process is sketchy [04:48:45] and making MW and all the deps into a farm of docker containers is a lot of coordination [04:48:59] but I'd love to be proven wrong [04:49:27] I had mw-vagrant "working" with a docker container as the target at one point [04:49:58] but it had a lot of issues due to the way I was treating a container as a VM [04:51:38] bd808: docker sure is interesting [04:51:59] the thing that I'm most concerned with is persistence with extensions [04:52:15] installing, updating, upgrading MW, etc. [04:52:51] wordpress on the other hand just persists its extension dir and manages it itself [04:53:09] MW doesn't have that capability [04:58:24] bd808: know anyone that can move that task along? [04:59:47] Negative24: I think if you add some description of a concrete goal I can probably talk YuviPanda into agreeing that it's worthy [05:00:07] will do [05:00:10] I'd actually be interested in helping at least as an advisor [05:00:35] We might be able to trick some releng folks into being interested too [05:01:22] I was first going to start with getting a stable MW version linked with an external db container and then work on persistence with extensions and such [05:02:24] what would need to be persistent about extensions? Just which ones were installed? [05:02:46] if you are upgrading MW you should almost always also update extensions [05:03:06] but anyhow yeah, write it up and I'll help sell it if it makes sense [05:04:37] K [05:18:18] (03CR) 10Sitic: [C: 032 V: 032] Set fontsize to 13px, fix minor bugs [labs/tools/crosswatch] - 10https://gerrit.wikimedia.org/r/229408 (owner: 10Sitic) [05:18:31] (03PS1) 10Sitic: Set uselang in MW API for diffs and notifcations [labs/tools/crosswatch] - 10https://gerrit.wikimedia.org/r/229627 [05:20:15] (03PS2) 10Sitic: Set uselang in MW API for diffs and notifications [labs/tools/crosswatch] - 10https://gerrit.wikimedia.org/r/229627 [05:21:58] (03CR) 10Sitic: [C: 032 V: 032] Set uselang in MW API for diffs and notifications [labs/tools/crosswatch] - 10https://gerrit.wikimedia.org/r/229627 (owner: 10Sitic) [05:22:09] legoktm: ^ fixed that language bug for notifications [05:22:43] \o/ [05:23:22] 6Labs: Labs Docker Project - https://phabricator.wikimedia.org/T106969#1513674 (10bd808) Suggested project goals: # Develop, test and publish Docker configuration for running a full stack MediaWiki install including MediaWiki, Parsoid and VisualEditor. # Stack should be able to link with a Docker container hosti... [06:06:50] 6Labs: Labs Docker Project - https://phabricator.wikimedia.org/T106969#1513691 (10yuvipanda) ^ sounds great :) if you are interested in this I can create the project as mediawiki-docker [06:10:51] 6Labs: Labs Docker Project - https://phabricator.wikimedia.org/T106969#1513703 (10Negative24) >>! In T106969#1513691, @yuvipanda wrote: > ^ sounds great :) if you are interested in this I can create the project as mediawiki-docker Sounds good. Docker seems a bit too broad. [06:12:13] 6Labs: Labs Docker Project - https://phabricator.wikimedia.org/T106969#1513704 (10Negative24) [06:13:15] 6Labs: Labs Docker Project - https://phabricator.wikimedia.org/T106969#1483024 (10Negative24) [09:49:00] 6Labs, 7Tracking: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514011 (10Green_Giant) 3NEW [10:43:22] valhallasw`cloud: hey [10:43:47] valhallasw`cloud: https://gerrit.wikimedia.org/r/#/c/164386/ is on my review queue -- I think the issue preventing this from being merged was fixed relatively recently by you [10:44:31] valhallasw`cloud: I just added you as a reviewer, feel free to amend/fix/push :) [10:48:54] paravoid: yes, that should be OK (https://gerrit.wikimedia.org/r/#/c/205914/). [10:49:12] right, that's the one I was remembering [10:50:57] the commit is essentially orphaned -- I doubt mark is going to work on it anytime soon :) [10:51:17] so it would be neat if you assumed control of it, test it and push it forward ;) [10:51:25] paravoid: is it possible to do a 'is a member of toollabs' check in exim ldap? [10:52:42] the advantage of a shell script is that it's a seperate thing that can be tested in isolation [11:01:28] 6Labs: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514177 (10Aklapper) [11:01:54] 6Labs: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514011 (10Aklapper) [11:27:28] valhallasw`cloud: I think not relying on a shell script is better, but feel free to -1/-2 the changeset if you disagree :) [11:40:59] valhallasw`cloud: it's easy enough to test ldap queries within the exim config as well [12:51:44] mark: oh, right, we can just hand exim a list of email addresses with the question 'where will you route this to' [12:51:56] yes [14:10:15] 6Labs: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514528 (10Bugreporter) Comment: # Some extensions are not needed, like GWToolset. others includes all CentralAuth/global* extensions, CheckUser, MassMessage, OAIRepository, WikiHiero, all Wikibase extensions, CentralNotice, NewUserMessage, SecureP... [14:11:41] 6Labs: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514534 (10Bugreporter) You can ask zhuyifei1999 for how to configure a new Commons-like wiki [14:14:06] 6Labs: NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514545 (10Bugreporter) Also you need another [[ https://www.mediawiki.org/wiki/Manual:Wiki_family | MediaWiki installation ]] as a client [14:15:29] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514549 (10Bugreporter) [14:19:47] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514551 (10Andrew) From the Labs terms of use: 4. Proprietary content: Do not use or create content unless it complies with the Wikimedia Licensing policy. This includes content in the public domain or freely licensed under a... [14:23:55] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514560 (10Bugreporter) @Green_Giant So the project should only use example images for demo. It should not host any real non-free files until the site is moved to production. [14:25:25] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1514562 (10Bugreporter) Or probably create a EDP for WMFlabs? [15:07:14] 6Labs, 7Tracking: New Labs project requests (Tracking) - https://phabricator.wikimedia.org/T76375#1514616 (10bd808) [15:12:16] bd808: thankya much [15:12:35] Negative24: go forth and make cool things for the world :) [15:12:43] I shall! [15:14:48] are we trying to go debian on everything now? [15:16:38] jessie is the preferred base image by techops now. The MediaWiki servers in production are still trusty because they were reimaged just before the switch to jessie and no one has had the energy to rebuild all 400+ or them [15:17:22] we don't have a custom hhvm build for jessie yet which to me is a blocker to migration of mw-vagrant [15:17:27] bd808: I don't know jessie as much so I'll probably still go with trusty but it shouldn't be that hard to figure out [15:17:56] granted that I've already had to muck my way through the sparsely documented openSUSE on my own [15:17:58] yeah, I haven't climbed the systemd learning curve yet either [15:18:18] ^ been learning it for docker + CoreOS [15:18:25] suse! I ran that back in the olden days [15:18:49] probably around '99 [15:19:04] * bd808 just outed himself as old again [15:19:06] YaST is love [15:19:14] YaST is life [15:19:39] I ditched it for ubuntu because of rpm hell [15:19:46] heh [15:20:08] I ditched ubuntu in favor for it because of unity hell and then went back to xubuntu [15:20:38] ah. I gave up on linux as a desktop around '04 [15:20:45] probably time to try again [15:21:42] what do you do work on? surely not windows. [15:26:25] OS X [15:26:40] which I use mostly as a nice window manager [15:27:12] my real work happens in tmux, vim, ssh, vagrant, etc [15:27:37] That ^ [15:28:20] In the cloud nobody knows that you are an Apple hardware fanboy [15:28:24] !bash [OSX] which I use mostly as a nice window manager [15:28:35] Aww, did I not do it right? [15:28:46] I think you did [15:28:51] jsut no ack from the bot [15:29:07] https://tools.wmflabs.org/bash/quip/AU8DoKfT6snAnmqnLGfa [15:29:19] Ah yerp [15:29:21] Gotcha [15:29:55] adding an ack might be nice... I wonder how immensely hard that would be? [15:30:16] Heh, make it say ack and give the url to the quip. [15:30:31] It would probably take another nick because of the way that logstash splits input and output [15:31:01] Boo [15:31:34] bd808: Totally unrelated. Mike made his paella yesterday. https://instagram.com/p/6Bcq-suKXs/ [15:31:36] #foodporn [15:31:56] yum [15:32:28] that looks ridiculously good [15:32:46] Mike and ^d eat well [15:32:50] the jerks [15:32:50] The pan was empty, heh. [15:33:25] +1 [15:34:10] * Negative24 forgot about the irc conversation [15:34:35] ostriches: hidden !bash trick: put in \t where you want to see a \n in the quip [15:34:45] I need to write up some docs soonish [15:34:49] ah ok [15:35:42] I'm probably also going to need to figure out how to do suppression/deletion as soon as people actually find it [15:35:57] whats this !bash thing? [15:36:05] Replacement for BZ quips [15:36:07] Negative24: https://tools.wmflabs.org/bash/random [15:36:35] I have a logstash + elasticsearch cluster that powers that [15:36:36] ha lol [15:37:02] that's amazing [15:37:22] it idles here and a few other channels as stashbot [15:37:37] heh, https://tools.wmflabs.org/bash/quip/AU7VUE6a6snAnmqnK_rk [15:37:50] then a php app running on tool labs to display & search [15:38:33] https://tools.wmflabs.org/bash/search?q=^d|demon|chad [15:39:25] ^d searches badly. seems to match on any "d" [15:41:21] this one almost made me spit out my coffee [15:41:23] what does +1 mean anyway? Reviewer has a working mouse. [15:41:34] bd808: Yep, that's totally why I picked my new nick. SEO. [15:41:52] and Joust! ? [15:42:40] https://tools.wmflabs.org/bash/quip/AU7VUnxR6snAnmqnK_sG ha [15:43:39] and now that we have introduced that time suck... back to email queue [15:43:57] same [15:47:50] YuviPanda: how hard is it to make another openstack image? [16:07:22] bd808: I want this to not be a problem on my system so could you try to ssh into docker-dev.mediawiki-docker.eqiad.wmflabs [16:08:42] Negative24: .... trying ... but getting nowhere [16:09:13] yep just giving me permanently_drop_suid [16:09:43] I just redid my ssh configs and I didn't want to recreate it and find it was my fault [16:09:52] I don't think the initial puppet runs are quite done yet -- https://wikitech.wikimedia.org/w/index.php?title=Special:NovaInstance&action=consoleoutput&project=mediawiki-docker&instanceid=b1cdbff4-78e6-4502-95c3-35e5a9d3c375®ion=eqiad [16:10:51] hmm I always get caught off guard with how long it takes [16:11:13] console doesn't seem to be changing from the last time I checked [16:11:37] it's not pinging from bastion-01 [16:12:26] probably just another fluke in instance creation. It'd be nice if those were fixed [16:12:56] 6Labs, 10Tool-Labs, 3Labs-Sprint-108: /srv on tools-static-01 running out of inodes - https://phabricator.wikimedia.org/T107657#1514820 (10Andrew) 5Open>3Resolved a:3Andrew [16:37:17] bd808: so supposedly the project came without ssh allowed [16:37:33] ummm.. huh [16:37:43] bad default security group? [16:38:13] bd808: seems like it. didn't have the source group nor icmp [16:39:49] good find. might be worth asking andrewbogott if that is a known bug/fluke that happens or a new badness [16:41:25] every time I create a new project and check, it has ssh access. But there must be a race because now and then I see it having not happeneed in retrospect [16:41:36] Anyway… is the immediate issue sorted out? [16:42:52] andrewbogott: yep [16:43:09] well I believe. I'm waiting for the new instance to spin up [16:44:03] yep all sorted out [17:05:23] !log stashbot Fixed conflicts with cherry-picked Puppet patches [17:05:59] labs-morebots: y u no log? [17:05:59] I am a logbot running on tools-exec-1214. [17:05:59] Messages are logged to wikitech.wikimedia.org/wiki/Server_Admin_Log. [17:05:59] To log a message, type !log . [17:08:00] !log tools.morebots dumdiedum? [17:08:05] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.morebots/SAL, Master [17:08:16] !log thisprojectdoesn'texist errr? [17:08:16] thisprojectdoesn'texist is not a valid project. [17:08:36] * valhallasw`cloud frowns [17:08:46] !log stashbot Fixed conflicts with cherry-picked Puppet patches [17:09:04] maybe it just hates me [17:09:32] !log stashbot sup [17:09:43] !log stashbot sup [17:09:53] ok, ssh'ing in [17:09:58] clearing the page didn't help either [17:10:18] UnboundLocalError: local variable 'header_date' referenced before assignment [17:10:22] yeah, something in parsing headrs [17:10:43] oh... it's trying to find the place to start a new date section [17:11:04] !log stashbot bleh [17:11:05] so blank page sadness [17:11:26] needs more ='s? :/ [17:11:55] I really don't get this [17:11:58] should be h2? [17:11:59] I thought I fixed this issue [17:12:32] wtf?! [17:12:37] !log stashbot Fixed conflicts with cherry-picked Puppet patches [17:12:41] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Stashbot/SAL, Master [17:12:52] so adding == 2015-08-06 == worked [17:13:06] I think I was part of making that change :/ [17:13:11] I guess tools-bastion has an old version maybe [17:13:15] so I'm looking at the wrong source [17:14:02] yeah, it breaks if there's no header. anything that starts with enough ===='s is fine [17:14:20] ugh [17:15:02] valhallasw`cloud: feel free to wipe and use https://wikitech.wikimedia.org/wiki/Nova_Resource:Stashbot/SAL for more testing if you have time to fix [17:15:11] bd808: it's a one-line fix, really [17:31:51] (03CR) 10Sniedzielski: [C: 032] Switched to new server and new folder structure [labs/tools/wikipedia-android-builds] - 10https://gerrit.wikimedia.org/r/229263 (owner: 10BearND) [17:34:31] (03CR) 10Sniedzielski: [V: 032] Switched to new server and new folder structure [labs/tools/wikipedia-android-builds] - 10https://gerrit.wikimedia.org/r/229263 (owner: 10BearND) [17:41:01] Hi. Does anyone know which table on the Labs DB contains the page view counts for the WMF wikis? [17:41:44] huji: https://dumps.wikimedia.org/other/pagecounts-raw/ [17:42:22] valhallasw`cloud: thanks! Is that also accessible on Labs DB so I can join it with a WP table? [17:42:30] not as far as I know [17:43:03] huji: WMF is working on making it more easily available, but it is a challenge due to the large amount of data [17:43:18] Nettrom: makes sense. [17:43:19] if you're looking for view stats on a daily basis for a limited number of articles, stats.grok.se is what most people use [17:44:04] Nettrom: I intend to tabulate the most viewed articles on a wiki, but redo it every week [17:45:41] Nettrom: but I guess I cannot do it fully programmatically. [17:46:57] huji: you might want to talk to User:West.andrew.g, looks like he maintains the weekly stats that the Signpost uses in its traffic report [17:47:26] Nettrom: that's very helpful! Thanks agian :) [17:47:41] huji: see https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2015-08-05/Traffic_report and https://en.wikipedia.org/wiki/User:West.andrew.g/Popular_pages (the latter is a huge page) [17:47:46] huji: np, glad I could help :) [17:48:53] 6Labs, 10Tool-Labs: Please install hugin-tools and pillow - https://phabricator.wikimedia.org/T108210#1515302 (10dschwen) 3NEW [17:49:54] 6Labs, 10Tool-Labs: Please install hugin-tools and pillow - https://phabricator.wikimedia.org/T108210#1515317 (10dschwen) [17:52:29] one more question: I have a bot that generates a very long page (1-2 megabytes) and fails to save the page on WP (gets a 503 error). Are there limits to the size of edits made by bot (or via API)? And what solution do you recommend? [17:52:38] the question is directed to everyone :) [17:52:49] huji: I've had the same issue with SuggestBot [17:53:07] (Suggestbot uses the pywikibot framework, btw) [17:53:23] even though you get a 503 error, the page might have been saved [17:53:28] Nettrom: good to know I am not alone :) [17:53:53] actually, it doesn't save. it just retries, and on every retry the delay is increased by 5 second. One time I left it for a minute but nothing happened [17:53:54] therefore, SuggestBot waits 60 seconds after an error like that, then checks the page's edit history [17:54:11] so you think I should elave it for longer? [17:55:21] huji: it depends... does the wiki actually not save the page? [17:55:35] Nettrom: well I stopped the bot after 1 minute [17:56:47] Nettrom: retrying now, won't stop it this time [17:57:02] huji: there's a slight difference in what the bot does and what the wiki does... in my case the page often actually gets saved, but the bot does not get the right confirmation [17:57:15] leaving the bot thinking the edit did not get saved, and it retries [17:57:31] hmm [17:57:34] interesting [17:58:48] ok, I got my first 503 .. leaving it for a few minutes [17:59:38] huji: I manually check the page's edit history [17:59:53] that way you'll know if the page actually got saved or not :) [18:00:55] huji: do you use pywikibot as well? [18:01:05] Nettrom: yes pwb [18:01:07] core [18:01:15] Nettrom: so for no edits have been made, per histor [18:01:35] huji: ah, then it's a slightly different problem [18:01:44] I have a meeting, unfortunately, so I gotta run [18:01:49] Nettrom: no worries [18:01:56] thanks for the help ): [18:01:58] :) [18:01:58] one thing I looked into when debugging this was parsing time of the page [18:02:02] sorry for the wrong emoticon :/ [18:02:23] if you're using lots of templates, you might want to look into how long they take to parse [18:02:42] Nettrom: it uses a lot of {{formatnum's [18:02:46] maybe that's the prob [18:03:06] huji: hm, maybe, but that's a parser function... I'd still look into it, though [18:03:16] (I'd expect parser functions to be fast) [18:03:29] anyways, I gotta run... back later, and good luck! [18:03:38] I wonder if there is alimit to how many parser functions can be run [18:03:44] i'll talk to you later Nettrom [18:37:50] Coren, hi. Any updates with the dedicated exec node? [18:43:24] If I create an instance web proxy how long should it take for DNS to propagate? [18:44:58] notconfusing: should be instantaneous I think [18:46:17] valhallasw`cloud, I verified I am serving on the backend, and I created the webproxy [18:46:33] notconfusing: I actually thought there was a catchall for wmflabs.org, but that's not true [18:46:43] notconfusing: are the security groups set correctly? [18:46:44] but i just get a timeout on http://wigi.wmflabs.org/ [18:47:03] Name: wigi.wmflabs.org [18:47:03] Address: 208.80.155.156 [18:47:05] valhallasw`cloud, the security groups on the web content? [18:47:09] yeah [18:47:15] try telnetting the webserver from bastion [18:47:22] ok [18:47:31] should it be 755? [18:48:08] notconfusing: eh, no, the security groups for the instance [18:48:16] basically the firewall rules [18:48:19] ah [18:49:31] ok, well the firewall is not open on port 80 [18:51:17] what rule do I want to add? port 80, tcp, right? what do I want in CIDR? [18:52:13] 10.0.0.0/32 [18:54:18] valhallasw`cloud, ok i added that rule, do i need to restart anything? [18:54:29] notconfusing: no, should work immediately [18:55:21] valhallasw`cloud, still getting 504 Gateway Time-out [18:56:49] notconfusing: have you tested from bastion? [18:59:24] ok, i'm telnetting from bastion [19:02:56] valhallasw`cloud, maximilianklein@bastion-01:~$ telnet wigi.wmflabs.org 80 [19:02:56] Trying 10.68.16.65... [19:02:56] Connected to wigi.wmflabs.org. [19:02:56] Escape character is '^]'. [19:02:56] GET / HTTP/1.1 [19:02:57] host: wigi.wmflabs.org [19:02:58] Connection closed by foreign host. [19:03:14] notconfusing: to the internal ip [19:03:21] wigi.wmflabs.org is the proxy [19:03:37] valhallasw`cloud, oh right [19:08:17] valhallasw`cloud, ok, I switched to CIDR 0.0.0.0/0 and now I get apache Bad Request errors, which is an improvement [19:08:24] from telnet from bastion [19:08:39] valhallasw`cloud, ah it completely works now! [19:08:48] * notconfusing appreciates valhallasw`cloud [19:08:59] * valhallasw`cloud is confused by the CIDR range then [19:11:35] valhallasw`cloud: yah, you needed 10.0.0.0/8 [19:11:41] ahahahahaha [19:11:44] /32 refers to a specific IP [19:11:45] yes, of course [19:14:11] YuviPanda, this docmentation says to use /0 which I guess is all IPs [19:14:13] https://wikitech.wikimedia.org/w/index.php?title=Help:Security_groups&redirect=no#Individual_rule [19:14:20] yup [19:14:22] that's ok too [19:16:45] YuviPanda: I'm writing unittests for tools-mail \o/ [19:16:51] woah [19:17:01] such dedication :D [19:17:06] thank you [19:17:10] exim has a 'fake smtp' mode [19:17:25] exim - and exim will respons as if you are [19:18:05] nice [19:19:22] and -bt to test routing [19:22:04] YuviPanda: I hope it's OK I also use your mail address for the test cases =[ [19:22:09] =P, I mean [19:22:25] valhallasw`cloud: :D sure [19:22:27] self.assertRoutesTo('tools.admin', 'yuvipanda@gmail.com') [19:22:34] I'm sure gmail can handle it [19:22:48] it doesn't actually mail [19:22:50] that's the best part [19:22:56] it's a completely offline test [20:22:54] 6Labs, 10Labs-Infrastructure: Switch to a multi_host nova network - https://phabricator.wikimedia.org/T107731#1515919 (10Andrew) [20:42:29] "Writing logs to NFS is bad m'kay. :)" in deployment-bastion:/data/project/logs . [20:43:24] but you didn't update https://wikitech.wikimedia.org/wiki/Logs#Beta_cluster , https://wikitech.wikimedia.org/wiki/Nova_Resource:Deployment-prep/Overview , etc. [20:43:34] I'll update ;-) [21:34:24] !log wikilabels created project, added madhuvishy and halfak as admins [21:35:25] bd808: is stashbot collecting these now? [21:35:25] morebots is dead again [21:35:31] !log wikilabels added wikilabels-01 instance, puppet group [21:35:44] YuviPanda: yes, it is [21:35:48] bd808: cool then [21:36:01] bd808: is there a place to see them? also do you need any puppet roles merged for stashbot? [21:36:06] and morebots may not be down so much as broken for empty pages [21:36:18] YuviPanda: https://stashbot.wmflabs.org/#/dashboard/elasticsearch/SAL [21:36:46] I have a role to merge but not until next tuesday when I update the prod logstash cluster [21:37:02] It was my first test case for logstash 1.5.3 [21:37:09] ah nice [21:37:10] ok [21:39:40] YuviPanda: valhallasw`cloud debugged the morebots thing for me this morning. It needs a small patch to not throw an exception when trying to add to a page that has no === headings === yet [21:39:57] fallout from the fancier section stuff [21:40:00] there's a patch in gerrit [21:40:38] and then the really stupid rebuildpackage dance [21:40:47] oh yeah :((( [21:41:12] I can do that, I think [21:41:16] except the package is on apt.wm.o [21:41:18] hm. [21:44:25] YuviPanda: if you fancy reviewing, https://gerrit.wikimedia.org/r/#/projects/operations/debs/adminbot,dashboards/default [21:44:48] !bash YuviPanda is literally labs-vagrant [21:44:53] I guess that took [21:45:30] anyhow, bed [21:45:36] oh, wait [21:45:43] let me first upload my exim stuff somewhere [21:46:49] YuviPanda: heh. labs-vagrant. I get it [21:48:36] valhallasw`cloud: merged 2/3 [21:49:23] valhallasw`cloud: please add your name [21:49:33] mutante: oops [21:49:52] eh, i merged but shouldnt have [21:49:59] it needs that line with the timestamp [21:51:39] https://gerrit.wikimedia.org/r/#/c/229970/ [21:51:52] mutante: thanks [21:55:06] YuviPanda: https://github.com/valhallasw/mailrelay-tests [21:55:33] (now I'm really off to bed) [21:58:17] YuviPanda: if i import it into repo, can you do the upgrade/restart? [21:58:24] it would be for precise [21:59:20] mutante: not now sorry, in a meeting [21:59:55] then i won't import it, i remember from last time there was "ensure latest" somewhere [21:59:58] ok [22:00:08] it's already built though and in the incoming dir on APT repo [22:01:23] ok [22:11:57] !log wikilabels recreated u_wikilabels user and db [22:12:13] !log wikilabels initialized server and db using fabric [22:27:03] !log wikilabels restarted uwsgi on server. created web proxy. labels-new.wmflabs.org is up now [22:27:10] halfak: ^ [22:27:13] doing data migration next [22:58:36] !log wikilabels migrated db from old wikilabels instance to labs pg store [23:26:04] bd808: how do I link to a specific search in kibana? [23:26:17] ah [23:26:17] https://stashbot.wmflabs.org/#dashboard/temp/AU8FVh036snAnmqnLHKx [23:26:29] 'share' [23:26:34] *nod* [23:26:52] or make a permanent dashboard for it with the floppy disk icon [23:27:01] a share lasts for 30 days I think [23:27:29] ah cool [23:27:55] !log wikilabels deleted labels proxy from old project, created in project wikilabels [23:28:34] !log wikilabels deleted temp labels-new proxy from project wikilabels [23:35:29] hello YuviPanda [23:35:37] hello leila [23:36:15] I'd like to add something at the end of the survey to let people know where they can contact the team, YuviPanda. Should this be: on IRC #wikimedia-labs or our email list? [23:36:23] leila: I'd say email yeah. [23:36:24] which email list should I provide, YuviPanda [23:36:52] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1516750 (10Green_Giant) >>! In T108167#1514528, @Bugreporter wrote: > Comment: > # Some extensions are not needed, like GWToolset. others includes all CentralAuth/global* extensions, CheckUser, MassMessage, OAIRepository, Wi... [23:37:06] what mailing list YuviPanda? [23:37:33] leila: labs-l I think? [23:37:38] leila: they can also mail me directly if they wish [23:37:55] Also, are you planning to share the raw data publicly? or you are planning to share the results of the survey publicly? we need to specify this at the beginning of the survey. [23:37:59] ooki [23:38:12] 6Labs: Create labs project for NonFreeWiki - https://phabricator.wikimedia.org/T108167#1516753 (10Green_Giant) >>! In T108167#1514528, @Bugreporter wrote: > Comment: > # Some extensions are not needed, like GWToolset. others includes all CentralAuth/global* extensions, CheckUser, MassMessage, OAIRepository, Wi... [23:38:37] leila: I think not [23:38:53] ooki, I'd say aggregated results will be shared? [23:39:02]