[06:27:43] hey guys, can i request deletion of my unused tools? [07:31:09] ebraminio: sure [07:33:42] !toolsdocs [07:33:42] https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help [07:38:53] @notify Coren [07:38:53] You've already asked me to watch this user [07:45:25] petan: should I request deletion of unused ones, to save resource? [07:45:43] you can, but I don't think you must [07:46:05] you should only request this in case you are never going to use them again [07:46:23] petan: how expensive are they? [07:46:29] I mean keeping those tools [07:46:36] basically they just clutter some pages [07:46:44] otherwise it's not a problem at all... [08:11:13] @notify Coren [08:11:14] This user is now online in #wikimedia-labs. I'll let you know when they show some activity (talk, etc.) [09:00:39] hi does anyone know what's wrong? how i can fix it? [09:00:44] http://tools.wmflabs.org/pywikibot/core/pwb.py [09:13:26] Amir1: it's running the script instead of displaying it [09:15:11] valhallasw: hi I know [09:15:28] I don't know how we can set it to showing not running [09:22:22] you can probably override it with .htaccess [09:22:57] yeah [09:47:35] petan: hi. are you still there? [09:48:25] yes [09:48:31] petan: all created by me tools [09:48:41] hmm [09:49:17] petan: fatestwiki (it has a db), transclusion, ebrambot, they are useless [09:49:55] ebraminio: I need a list of tools to delete, because 1. I don't know your wikitech name, 2. There is nothing like "creator of a tool" only members [09:50:24] Ebrahim ? [09:50:26] petan: honestly i couldn't figure out how I can use replications db from my tools [09:50:38] petan: yes, i emailed too root@.... also [09:50:40] well, did you read docs? [09:51:04] !toolsdocs [09:51:04] https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help [09:51:11] there is explanation on how to use replicated db's [09:51:59] petan: okay, so please just delete fatestwiki (and its db) and ebrambot [09:52:05] ok [09:52:40] !log tools petrb: deleting fatestwiki tool, requested by creator [09:52:42] Logged the message, Master [09:52:47] petan: can you help me in an issue? [09:53:31] petan: http://tools.wmflabs.org/pywikibot/core/pwb.py [09:53:48] what i have to write in .htaccess [09:57:11] AllowOverride None [09:57:13] Amir1 ^ [09:57:26] ah wait [09:57:57] Options -ExecCGI [10:01:01] petan: thank you [10:01:14] petan: I have not replica.my.cnf in my tool home so please remove "transclusion". Hope someone with better skill on wmflabs use the name. Thank you [10:01:37] *I hope [11:09:16] petan: Can you remained tools? I don't want waste wmf resources [11:09:23] *delete [11:21:41] sure [11:24:51] ebraminio anything else? [11:27:17] petan: Thank you. Do you have some sample configuration for setting up a wiki via puppet? Just wanted to learn if is possible (and no problem if is not) [11:27:30] no I don't think it's possible yet [11:34:31] petan: okay, thank you [11:40:48] petan: you around? [11:40:54] yes [11:41:20] do you think there are bots that depend on the session cookie format being 'foo_session' [11:41:27] and that would break if it were set to FooSession? [11:41:43] better yet, is the code for bots on some central instance that could be grepped through? [11:42:06] no, there is no such instance [11:42:47] hrm, there's https://wikitech.wikimedia.org/wiki/Nova_Resource:Bots (obsolete) and Tools [11:42:47] but I don't really know, it's is quite possible that some tool could be affected by this change, which IMHO is more cosmetical than practical :P [11:43:05] we're a cosmetical bunch [11:43:12] heh [11:43:25] I don't mind saying, there's nothing cosmetic about it [11:43:41] I think we'll just have to change it now and announce it afterwards [11:43:43] bots do live on tools project now, but some of them don't have source code there, only compiled binaries [11:44:33] ori-l: can you push this to beta.wmflabs.org first so that someone can test their bots there [11:44:49] probably not, it's a bit urgent [11:45:14] ori-l: for pywikipedia, it just uses a cookie jar which stores whatever the server sends. Not sure about other frameworks. [11:46:40] is there a mailing list? [11:46:44] good :) could you folks keep an eye open for weird bot authentication failures, and suggest that the old cookie name format may be hard-coded to anyone debugging? [11:46:51] ori-l: I think it wouldn't break any of my bots but I can't speak for everyone... [11:47:47] !mail [11:47:47] we have a mailing list labs-l@lists.wikimedia.org feel free to send a message there, don't forget to subscribe [11:47:51] TimStarling ^ [11:47:57] mailing the API list would also make sense [11:48:39] however: why is the change necessary in the first place? or will that be announced afterwards :p [11:52:29] for security [11:52:49] the details are probably best left until after the deployment, although plenty of hints have been dropped in public channels already [11:54:14] basically it is the simplest way to reset all session IDs [12:17:38] !toolshelp [12:18:04] !tools-help [12:18:04] https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help [12:18:16] !toolshelp is https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help [12:18:17] Key was added [12:18:43] !toolshelp del [12:18:44] Successfully removed toolshelp [12:18:45] http://bots.wmflabs.org/~wm-bot/dump/%23wikimedia-labs.htm [12:19:00] !toolshelp alias tooldocs [12:19:00] Created new alias for this key [12:19:04] !toolshelp [12:19:04] https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help [12:19:07] ah. That's better. [12:19:21] one day that url may change, we don't want to update 200 keys ;) [12:20:17] !tools-help del [12:20:18] Successfully removed tools-help [12:20:24] !tools-help alias tooldocs [12:20:24] Created new alias for this key [12:22:23] petan: can you reat a new function for wm-bot? [12:22:30] reat? [12:22:36] creat [12:22:39] ah [12:22:43] which one [12:23:11] blabla [[:de:test]] blaba [12:23:11] wm-bot: https://de.wikpedia.org/wiki/test [12:23:11] Hi Steinsplitter, there is some error, I am a stupid bot and I am not intelligent enough to hold a conversation with you :-) [12:23:32] we have such a funktion in the linky bot [12:23:36] hmm [12:23:49] but if we can inlucde dis in wm-bot, /kick linkybot [12:23:53] possibly [12:23:59] what else does your linkybot do? where can i see it? [12:24:14] but where do I get a db of all wiki prefixes [12:24:22] you don't need to [12:24:29] howcome [12:24:41] gry: feel free to join #wikipedia-de or #wikipedia-bio [12:24:47] [[:meta:bla]] [12:24:55] jepp, such this [12:24:58] how the bot knows it should go to meta.wikimedia.org [12:24:59] http://repo.or.cz/w/gpy.git/blob/HEAD:/lib/WikiLinkParser.pm [12:25:07] petan: ^ [12:25:29] 14:25:18 - Steinsplitter: [[:meta:bla]] [12:25:29] 14:25:18 - linky: http://meta.wikimedia.org/wiki/bla?redirect=no [12:26:20] gry: what language is that and what is it for? [12:26:31] seems to me like something what convert a wiki page to html [12:26:36] not exactly what I need [12:26:59] no, a [[]] to an http link [12:27:15] http://repo.or.cz/w/gpy.git/blob/HEAD:/WikiLinker.pm [12:28:22] gry: I fail to see how does it construct the domain in url [12:28:48] this is a [[test]] blah [12:28:48] http://de.wikipedia.org/wiki/test?redirect=no [12:28:53] server-side [12:28:57] this is a test [[:en:blah]] [12:28:58] http://en.wikipedia.org/wiki/blah?redirect=no [12:29:08] so it doesn't really construct such domain itself, no [12:29:10] there is no need [12:29:14] no [12:29:20] does it use api or something for that? [12:29:23] yes [12:29:38] [[:wikitech:hi]] [12:29:38] http://de.wikipedia.org/wiki/wikitech:hi?redirect=no [12:29:41] [[:blabla:test]] [12:29:41] http://de.wikipedia.org/wiki/blabla:test?redirect=no [12:29:56] Be quiet! :p [12:30:04] it should consider switching to https :-) [12:30:25] jepp. i dos not use it , onli .de [12:30:28] {{:n:W|hi}} [12:30:28] http://de.wikinews.org/wiki/W?redirect=no [12:30:29] http is evil and bad [12:30:41] {{evil}} [12:30:42] http://de.wikipedia.org/wiki/Template:evil?redirect=no [12:30:48] !toolserver2 [12:30:48] toolserver2 sagt mir nichts [12:30:50] i think that borked a bit [12:30:57] !toolserver [12:30:58] toolserver sagt mir nichts [12:31:04] ??? [12:31:12] O_O [12:31:23] !ts [12:31:23] ts sagt mir nichts [12:31:23] !hilfe [12:31:28] !ts2 [12:31:34] !logs [12:31:34] http://bots.wmflabs.org/~wm-bot/logs/%23wikimedia-labs/ [12:31:46] !dns wikipedia.de [12:31:47] https://labsconsole.wikimedia.org/wiki/Help:Addresses [12:31:47] ts2 sagt mir nichts [12:31:48] logs sagt mir nichts [12:31:50] wikipedia.de ist 195.10.208.211 [12:32:03] !logsearch [12:32:03] logsearch sagt mir nichts [12:32:03] http://bots.wmflabs.org/~wm-bot/searchlog [12:32:17] can we move this to -offtopic [12:32:18] @q Stonebot2 [12:32:44] Cyberpower678: i hav schowed a function to petan... [12:32:51] ? [12:33:01] ther is no reson to quit it if i schow petan how the bot work [12:33:03] wtf... [12:33:18] quit what? [12:33:26] @q Stonebot2- [12:33:27] Permission denied [12:33:30] @-q Stonebot2- [12:33:31] !logger [12:33:34] ..... [12:33:37] * nerus1 :P [12:33:42] @unq Stonebot2 [12:33:51] !stirb [12:33:58] !quit [12:34:01] /-( [12:34:21] man if you want it to part this channel there is a command for that [12:34:37] no... i want it as part for wm-bot [12:34:44] to use it in other channels [12:34:47] I am trying to improve this - http://tools.wmflabs.org/?status [12:34:55] nerus1: cool [12:35:19] Steinsplitter: ok I need to figure out how to do that api resolution in some simple way [12:35:35] so that it works smooth when wmf servers are down etc [12:35:55] what is preferred way to develop this - I was thinking may be create a new tool and put the new code there? [12:36:12] petan: ^^ [12:36:12] ah [12:36:15] !newlabs [12:36:15] This is labs. It's another version of toolserver. Because people wanted it just like toolserver, an effort was made to create an almost identical environment. Now users can enjoy replication, similar commands, and bear the burden of instabilities, just like Toolserver. [12:36:49] :p [12:37:09] @unq Stonebot2 [12:37:13] nerus: I don't really know... maybe you can create a new tool for that [12:37:26] Steinsplitter, Stonebot2 was in my way. [12:37:45] :P [12:38:18] petan: yeah. otherwise i would need access to tool.wmflabs.org's www and that would be too much ;) [12:38:35] of course i am assuming thats where the current status page is served from [12:38:35] I think you already have it [12:38:41] it's a git repository [12:38:53] everyone can submit patches to it [12:38:59] ah I see [12:39:23] !Coren [12:39:23] Coren is dead. petan killed him. He now roams about as a zombie. [12:39:38] petan, where's Coren [12:39:51] how could I know that [12:40:05] !Coren [12:40:05] Coren is dead. petan killed him. He now roams about as a zombie. [12:40:13] petan, ^ [12:40:22] /ignore Cyberpower678 [12:40:55] the problem is that I can't test stuff while developing. the data is a system call to qstat and i won't be able to do that in local machine. Unless I am missing something [12:41:19] nerus: you can develop on toolsbeta [12:41:25] petan, that wasn't serious was it? [12:42:05] Cyberpower678: that was just a typo, I wanted to execute it as a command [12:42:20] :-( [12:44:02] !log toolsbeta rebooting apache it seems to be frozen [12:44:04] Logged the message, Master [12:45:42] petan: oh cool! [12:46:24] !toolsbeta [12:46:31] !tools-beta [12:46:32] http://tools-beta.wmflabs.org/ [12:46:39] the webserver is rebooting now [12:47:05] !toolsbeta is development environment for tools project http://tools-beta.wmflabs.org/ [12:47:05] Key was added [12:47:24] thanks [12:47:47] it works now [12:47:53] http://tools-beta.wmflabs.org/?status [12:53:44] Is there any limit on how much space you can use in the shared storage space for a project on tool labs? [12:54:29] A volunteer wants to host a GiS project there, but says it will need 10 gigs of space for the data files. [12:54:35] petan: ^ [13:01:07] no I don't think there is a limit yet, but there will likely be a limit in future (I think), the shared storage has plenty of space now... [13:01:41] more than 13TB free [13:05:08] petan: i couldn't login to tools-beta though. i went ahead and created a tool. we could delete it later though [13:05:35] petan: we could delete it later though?* [13:06:05] yes I can delete it... [13:12:51] !ping [13:12:51] !pong [13:13:00] !ping [13:13:00] !pong [13:15:28] ping? [13:15:35] !poof [13:15:36] *POOF* "Wadda need?" *POOF* "Wadda need?" *POOF* "Wadda need?" [13:15:53] :O [13:17:07] petan: ? :P I just got am amusing message from wm-bot ;p [13:17:09] this url is probably not a valid rss, the feed will be disabled, until you re-enable it by typin..... [13:17:18] but the message ends there and doesnt tell me what to type :D [13:17:19] yes :P [13:17:25] what [13:17:27] where [13:17:33] im guessing the total line is too long [13:17:41] meh [13:17:52] you don' [13:17:59] t need to type anything it will not work anyway [13:18:08] I need to find or code some proper rss feed parser [13:18:12] so the feed wont ever work again? :P [13:18:18] the feed you are using is likely not atom [13:18:23] I doubt it ever worked to you [13:18:25] its bugzilla :P [13:18:28] oh lol [13:18:31] and has been working fine xD [13:18:31] that one should work [13:19:12]

DEBUG Exception in module Feed: a name did not start with a legal character 41 ()) Line 17, position 31. [13:19:19] what happened to http://bots.wmflabs.org/phpmyadmin/ ? [13:19:32] kaldari: it was disabled by wmf [13:19:50] petan: 3:12 PM <•wm-bot3> Unable to parse the feed from [13:19:50] https://bugzilla.wikimedia.org/buglist.cgi?%2Cassigned_to%2Cpriority%2Cbug_severity%2Cbug_status%2Cresolution%2Cshort_desc%2Cchangeddate%2Ckeywords%2Cvotes&email1=wikidata-bugs%40lists.wikimedia.org&emailassigned_to1=1&emailcc1=1&emailtype1=substring&query_format=advanced&resolution=---&resolution=LATER&ctype=atom&list_id=226873 this url is probably not a [13:19:50] valid rss, the feed will be disabled, until you re-enable it by typin :< [13:19:51] is there a new phpmyadmin interface somewhere else? [13:19:55] because it wasn't secure enough for them, some say that phpmyadmin is full of security holes [13:20:16] kaldari: there is no phpmyadmin, because wmf prohibited usage of phpmyadmin anywhere on wmf servers [13:20:59] addshore: you just need to type @rss+ [13:21:06] kk :> [13:21:07] to re-enable it but it will break again [13:21:17] oh wait, rss+ and RSS+ are different arent they? O_O [13:21:19] because the bot has tried 2 or more times before it gave up [13:21:22] idk [13:21:25] hah! [13:21:27] @RSS+ blah [13:21:33] I think upper case doesn't exist [13:21:37] @rss+ blah [13:21:38] There is no such item, if you want to define new item, please use 2 parameters [13:21:42] yup [13:21:49] petan: so there is no phpmyadmin interface at all now? [13:21:55] kaldari: correct [13:22:54] any idea who actually is enforcing this decision? [13:32:39] kaldari: Coren, Ryan and others.... [13:56:35] kaldari: http://www.mail-archive.com/labs-l@lists.wikimedia.org/msg00895.html [13:57:05] thanks [13:58:38] I don't necessarily need it myself, but it seems I can imagine fewer people will want to migrate without it [14:02:49] at the end of that thread matt recommends mysql workbench with built-in ssh tunneling, so people could probably run that locally and connect to labs. maybe one of Coren petan andrewbogott_afk could confirm that that's okay, if it is. [14:03:12] it is technically possible, but complicated like hell [14:03:38] I don't recommend it to anyone :o unless they are UNIX guru's and these are fine with terminal sql client... [14:03:48] oh. [14:07:07] Silke_WMDE: Thanks for the suggestion though. Might be useful for some folks. It would be nice if there was a section in https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Database_access briefly discussing the issue, since it doesn't appear to be documented currently. [14:07:37] you're right [14:17:47] ah Silke_WMDE I was wrong I just figured out that using our current configuration, where tools-dev is directly accessible it's quite easy to setup workbench [14:18:04] when I last tried it I had to socket UDP tunnel through bastion [14:18:16] but because tools project has own bastions with direct access to sql it's different [14:37:37] petan: Oh, good. We should find someone who is using it and would like to add a howto to the docs. [14:42:43] we should probably set up a postgre instance :> it's quite better than mysql in many ways [14:45:46] a user like to help me moving https://toolserver.org/~luxo/contributions/contributions.php to wmftools? [14:48:07] Steinsplitter: what kind of help you need [14:48:38] the last time was a prblem with the crosswiki-db [14:49:40] is a hig-used tool :D [14:49:50] and toolserver is broken and Luxo is inactive [14:50:11] well you explained why you need to move it but I asked what kind of help you need? :D [14:50:29] with the database :D [14:50:30] 1. step is to create a tool [14:50:51] 2. set up a git repository for source code [14:50:57] 3. upload the source code to repository [14:51:15] 4. change the source code so that it works on labs [14:51:49] OMG. [14:52:01] * Steinsplitter dos not hav time to work with GIT [14:52:21] ... [14:52:29] no git account *duck* ... [16:32:46] [bz] (8NEW - created by: 2Chris McMahon, priority: 4High - 6enhancement) [Bug 50335] support dvwiki in beta labs - https://bugzilla.wikimedia.org/show_bug.cgi?id=50335 [17:15:32] kaldari: It's actually not that hard, really. It takes one bit of configuration in your ssh client is all.\ [17:15:43] kaldari: (workbench through ssh) [17:17:05] [bz] (8NEW - created by: 2Tim Landscheidt, priority: 4Unprioritized - 6normal) [Bug 53055] Block Laryomaima and Laryomaima234 - https://bugzilla.wikimedia.org/show_bug.cgi?id=53055 [17:17:42] BTW, the Toolserver doc re DB forwarding can be used 1:1. [17:18:10] scfc_de: I should expect it would; ssh is ssh after all. [17:28:15] Coren, petan: Whenever I try to run any PHP script from jsub on the tool labs server I get the following error: libgcc_s.so.1 must be installed for pthread_cancel to work. Any idea what that's about? [17:29:38] Hmm: "Encountering this while trying to run jobs on the grid engine means you need to give your job more memory" [17:30:18] how do I give it more memory? [17:31:31] Seems strange that it would need more memory to execute an echo statement [17:32:46] kaldari: With the -mem parameter to jsub. [17:32:58] looks like it's supposed to have 256MB by default [17:33:46] kaldari: PHP does have a large footprint to fire up, even though it's relatively frugal once started. IIRC, most people report that 320m is enough. [17:33:59] ah, thanks! [17:39:21] Coren: still can't get it to work :( [17:39:41] jsub -N test php /data/project/hotarticles/public_html/test.php -mem m350 [17:40:20] Unix style; options before parameters. You want jsub -N test -mem 350m php /data/project/hotarticles/public_html/test.php [17:40:30] oops [17:40:42] Ortherwise, you're giving "-mem m350" as an option to your test.php :-) [17:42:21] it works now :) [17:42:47] I'm going to add all this to the documentation [17:52:46] That fits in the "When you've been doing it for 25 years, it's obvious" category. :-) [17:53:58] Obvious... [18:35:53] [bz] (8NEW - created by: 2Chris McMahon, priority: 4Unprioritized - 6major) [Bug 53061] support Flow on beta cluster - https://bugzilla.wikimedia.org/show_bug.cgi?id=53061 [18:42:03] Coren: you saw my labs-l note re docs plans? [18:42:31] About pushing the draft live? Yes. Silent consent. :-) [18:44:40] OK :) [18:45:09] Coren: and do you think you could make time to do that giant sweep with me & Silke re old wiki pages? [18:46:41] sumanah: I can probably squeeze it in; thankfully much of the NFS stuff has necessary delays built into it so unless you need a strict schedule I'm here. [18:46:53] ah, got it [18:47:54] I want to fix this asap, while the current workaround is servicable, we lost a layer of redundancy and snapshots so normalizing this is at the top of my list. [19:05:48] !log deployment-prep rebuilding the search indecies to pick up some recent changes [19:05:52] Logged the message, Master [19:07:19] thank you, bot [20:53:51] Coren: how are you feeling? hope you get well & rested soon [20:54:43] sumanah: I'm at the tail end; that flu mostly ruined my weekend. :-) [20:55:18] The bigger problem is that I'm rather heavily medicated so I'm not all there. :-P [21:00:23] Coren / petan, could you install python-beautifulsoup on labs? [21:03:53] (please install from source? or is that not possible) [21:04:10] valhallasw: hi, how are you today? [21:04:26] valhallasw: I know kma500 & I would love to get your feedback on https://www.mediawiki.org/wiki/User:Kmenger/ToolLabsGuide [21:05:17] hi sumanah! [21:05:52] this is mainly to get pywikipedia working, for which the version is not that important [21:06:44] I'll take a look at that page tomorrow (hopefully) [21:06:45] valhallasw: Please open a bugzilla so it doesn't fall between the cracks. If nobody else gets to it, I'll probably be able to do it tomorrow. [21:07:31] ugh, which project is that... [21:07:41] oh, under labs, not under tools [21:08:04] valhallasw: Hold on :-). [21:08:43] [bz] (8NEW - created by: 2Merlijn van Deen, priority: 4Unprioritized - 6normal) [Bug 53072] Install python-beautifulsoup on Tool Labs - https://bugzilla.wikimedia.org/show_bug.cgi?id=53072 [21:08:49] ^ [21:09:39] Coren: thanks in advance :-) [21:09:45] * valhallasw has to go now [21:09:54] thanks valhallasw [21:10:01] woo [21:10:03] :) [21:10:37] * sumanah cheers a bit [21:11:06] Is it possible to have mysqlnd on the servers ? [21:11:13] hmmm, any updates on the NFS situation while I was *actually* away from the computers? [21:11:43] YuviPanda: now stable (because of workaround), Coren did it [21:11:52] Coren: https://gerrit.wikimedia.org/r/79921 [21:11:57] the cache increase? or was the underlying problem fixed? [21:13:15] YuviPanda: No, switch to the internal drives thus not using the problematic controller. [21:13:21] wooo! [21:13:23] YuviPanda: I.e.: not tenable at long term. [21:13:40] still RAID, right? [21:13:48] Yeah, but one less layer of protection. [21:14:01] (I.e.: straight raid6 rather than a raid6 of raid1s) [21:14:27] ah, right [21:14:34] Coren: did we at least isolate the problem? [21:14:40] (sorry if this was on labs-l, just landed) [21:14:59] YuviPanda: That comes this week; I'm going to do testing on the hardware now that the immediate issue is circumvented. [21:15:21] Coren: ah, sweet! Ty! [21:16:28] YuviPanda: Is the grrrit-daemon offline? [21:16:33] scfc_de: no? [21:16:46] scfc_de: shouldn't be, at least [21:17:25]