[00:05:59] hedonil, http://pastebin.com/KbZSPMPR [00:08:26] Cyberpower678: ok [00:08:42] Program bug? [00:09:39] Cyberpower678: If you have the possibility to get a frsh key-pair generated on linux system with ssh-keygen -t rsa we can proceed [00:10:10] Cyberpower678: right now I think it's a key problem [00:10:22] Isn't it possible with the labs interface? [00:10:54] Cyberpower678: for me without agent pagent.exe MySQL WB couldn't handle the putty-gen keys [00:11:25] Well. I'm going to eat dinner now, and then I'll be back. [00:11:33] Cyberpower678: MySQl WB cannot access the private key [00:11:47] The T100 gets up and leaves [01:23:13] Cyberpower678: ready to rumble ? [01:24:05] Working on it now. [01:25:12] Cyberpower678: Here my suggestion http://tools.wmflabs.org/paste/view/bffb14f8 [01:25:54] hedonil, I've already done 1-6 :p [01:26:05] Cyberpower678: yeah [01:27:48] Didn't work. [01:27:58] Cyberpower678: logs? [01:28:08] Let's try some more tomorrow. I'll give you the logs. [01:28:17] Cyberpower678: ok [01:30:07] hedonil, http://pastebin.com/s9H1jXvg [01:30:12] Good night [01:30:25] Cyberpower678: sleep well [01:31:03] Cyberpower678: pagent was still running [01:31:41] Hi. [01:31:48] Can anyone help me with threading in PHP? [02:28:08] Change on 12mediawiki a page Wikimedia Labs/Tool Labs/List of Toolserver Tools was modified, changed by Billinghurst link https://www.mediawiki.org/w/index.php?diff=868095 edit summary: link fix [02:28:26] hehe [04:44:53] Coren: something's up with tools-login [04:45:04] `become` takes tens of seconds [04:45:17] and it's definitely not a problem with my connection [04:47:28] Yeah, it is slow. [04:47:36] But it /does/ work. [04:54:20] not necessarily a tools-login problem, could be ldap i guess [04:54:33] but that should be cached to some extent i thought [04:54:46] seems intermittent [04:55:00] just took me 39 secs once [04:57:23] yeah, that time about the same amount of time [04:57:42] i'm thinking ldap's the bottleneck [04:58:08] now once i got 14 secs [04:58:29] re my uplink connection, there's an unrelated problem: http://dpaste.com/1533138/plain/ [04:58:51] locally i have no loss and no loss to the other end but intermediate hops are horrible [04:59:17] (not on one of my normal links atm) [04:59:45] "uplink connection", huh, i must be tired [05:42:43] anyone around? [05:43:50] no [05:43:55] hi Magog_the_ogre [05:44:05] (you probably shouldn't be asking that) [05:45:00] I just rsync'ed everything from my directory into the local- directory of my tool [05:45:14] but the web server can't access it due to lack of permissions [05:45:23] when I tried to chmod on the file, it just says "operation not permitted" [05:45:44] how the heck am I supposed to put my files into the tool directory if the tool can't read it, and I can't *allow* it to read it? [05:47:02] sounds like you want take [05:47:29] https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Using_.27take.27_to_transfer_ownership_of_uploaded_files [05:48:29] that was... far too easy [05:48:30] thank you [05:50:38] ... or it would be it had functioned properly [05:50:53] local-magog@tools-login:~$ ls log -l [05:50:54] ls: cannot access log/category-files-20131230004005.out: Permission denied [05:50:54] total 0 [05:50:54] -????????? ? ? ? ? ? category-files-20131230004005.out [05:51:00] who do I report this to? [05:52:32] what did you run exactly? [05:53:03] take * [05:53:36] i guess Coren? i'd look myself but it's time for sleep [05:54:02] OTOH, i share a TZ with Coren [05:54:22] yes [05:54:41] I have insomnia tonight myself (a combination of caffeine, a midday nap, and a lack of exercise today) [05:55:09] it's not critical it's done today [05:55:26] but the tool server will shut down on the 6th, and I'd like to have my tools running on labs by then [05:55:46] will it really? [05:55:53] i thought that's just an expiry [05:55:57] you could dig more yourself... [05:56:15] you also could work around it with cp [05:56:18] or rsync [05:56:34] well I COULD [05:56:59] except that now all the permissions are stuck in spiritual purgatory and NEITHER of my accounts appears to have access to them [05:57:40] you could delete all the stuff you copied and rerun the rsync [05:57:56] and then within the tool account run another copy to fix perms [05:58:02] no, I can't [05:58:10] you cant [05:58:13] you can* [05:58:19] good night :) [05:58:28] gosh darn it [05:58:44] ok, who do I ask that can run a sudo and delete the files tomorrow [05:58:44] ? [05:59:02] should I message Coren on the wiki? [05:59:59] I will send Coren a message on wiki [06:00:15] Magog_the_Ogre he's not active on-wiki though [06:00:18] try to memo him on IRC [06:04:49] ok, thanks all [06:05:20] pardon me if I sound grumpy; it's just frustrating when things don't work the way you expect them to. :) [08:21:05] (03PS1) 10Hashar: Jenkins job validation (DO NOT SUBMIT) [labs/toollabs] - 10https://gerrit.wikimedia.org/r/104917 [10:46:26] petan: Around? [10:46:40] hi [10:46:58] . [10:47:08] what's up [10:47:30] petan: Why am I not in the "wikidev" group, although most other users seem to be? [10:47:47] Currently helping a user and I guess it would be nice if I could peak into his $HOME [10:48:13] that is some ancient group that all users were members of when labs were launched [10:48:25] it's not being used for new users [10:49:02] that user can always give you access different way [10:49:08] they can even change the group [10:49:09] petan: mh... I know [10:49:43] But still :/ [10:49:59] I'm in the svn group even, but not there and I don't think that makes much sense [10:51:12] petan: That's clearly a disadvantage attached with my account... a bit of a mess actually [10:51:15] maybe it's the other way... new users are in wikidev and old were in svn [10:51:30] however this shouldn't matter to you [10:51:41] if you want to have access to files of that user you need to ask them [10:52:57] mh, still not really happy with that... if you don't want people to have access per default don't use shared groups at all [10:53:20] this isn't anything I set up neither anything I can change [10:53:28] you need to talk to Ryan if you want to fix this [10:55:48] Will (maybe) do [10:56:08] just thought this is a messed that should be cleared up... thanks for your help anyway :) [10:56:38] I hope I can get my tools (and APIs) moved over this month... they mostly output crap on TS these days [13:10:50] hedonil, ping [13:15:33] petan: why ryan? [13:15:58] jeremyb: because Ryan is LDAP master [13:16:15] and labs is his baby [13:16:23] huh? [13:19:36] Cyberpower678: Hi CP, any news from the culprit? [13:19:46] No [13:19:53] Sadly. I just got up. [13:20:31] Cyberpower678: yesterday I just saw in the logs, that pagent with the old key was still running. It has to be terminated with the new keys [13:22:01] Cyberpower678: otherwise pagent takes over control everytime a ssh connection is build up [13:22:20] Connection established [13:22:43] Cyberpower678: So you are in? [13:22:50] Yes [13:22:59] Yay [13:23:15] Cyberpower678: \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ [13:25:18] * hedonil is going to write this workaround to the docs [13:25:29] That's a good idea. [13:52:00] hashar: beta seems unhappy [13:53:00] en.wikpedia.beta [13:54:12] simple wikipedia is ok. I'll continue testing there for the time being [13:54:21] !log deployment-prep upgrading Elasticsearch servers in beta [13:54:46] Logged the message, Master [13:58:14] (03PS1) 10Tim Landscheidt: Fix build and run-time dependencies [labs/toollabs] - 10https://gerrit.wikimedia.org/r/104951 [14:05:38] petan, ping [14:06:00] manybubbles: yeah message files on beta are no more updated. Fix is https://gerrit.wikimedia.org/r/#/c/104953/ [14:06:47] Coren|Travel, ping [14:07:33] !log deployment-prep running mw-update-l10n , it was broken because of https://gerrit.wikimedia.org/r/#/c/104741/ fixed up by https://gerrit.wikimedia.org/r/#/c/104953/ [14:07:46] Logged the message, Master [14:07:56] hashar: thanks! [14:08:05] anomie, are you a root on tool labs? [14:08:11] Cyberpower678: no [14:08:42] Is somebody able to purge tools-login? [14:09:12] Someone decided to run a heavy task on it again and it takes about a minute for me to access tools. [14:09:49] Tanvir: ping [14:10:39] Hello hedonil. [14:10:42] Tanvir: Hi, you are running some heavy tasks on tools-login [14:10:47] Tanvir: nano [14:10:54] Okay. [14:11:04] Which one do you mean? [14:11:21] Tanvir: I can only see pid 4207 [14:11:30] Tanvir, are you running any task on -login? [14:11:41] Because that's a no-no [14:11:42] Tanvir: and it is blocking tools-login [14:11:55] Wait a minute. [14:12:00] Tanvir: It looks like you have an editor open for your crontab, and it's burning CPU for some reason. [14:12:28] If you could close it or kill it, that would be helpful. [14:12:31] pkill -u local-tanvir [14:12:41] Done. [14:13:21] Tanvir: fine ;) you can login to tools-dev.wmflabs.org and run your heavy task there [14:13:38] That didn't seem to help it. [14:13:49] Tanvir: It appears to still be running. Cyberpower678 doesn't know what he's talking about, BTW [14:14:17] anomie, it's what I use to kill stay tasks on login. [14:14:42] Tanvir: If you don't have anythink open to edit your crontab that you know of, try the command "kill -9 4207" when logged in as your user account (not your tool account) [14:15:27] Trying that. [14:16:23] Did it help? [14:16:29] No [14:16:32] Tanvir: The problematic process is gone now. Cyberpower678, did it help? [14:16:39] Tanvir: yes [14:16:45] Coren|Travel: something's up with sudo in labs. not just in tools [14:16:48] takes way too long [14:16:49] Going down now [14:16:52] * anomie sees nothing much using CPU on tools-login now [14:17:04] Good to know, anomie. [14:17:06] anomie: He killed the nano process [14:17:13] crontab and nano again? :/ [14:18:00] ah. that apparently happens if you run nano, and your ssh connection dies [14:18:45] * Cyberpower678 needs to study linux. [14:18:48] :p [14:19:26] but I can't reproduce it locally [14:20:10] didn't we have a cpu time killer tools-login or something like hta [14:20:33] I had stuff killed over there than I did data processing (I know that's a no no) [14:20:51] hoo: tool is named Coren ;) [14:22:21] I think it was automatic, cause my mysql client ate to much ram [14:25:50] Cyberpower678: linux lesson no 1: ls (list files), cat (show file content), top (monitor system resources), exit (leave) :P [14:28:19] :D [14:28:25] Thanks. [14:28:46] Cyberpower678: you're welcome ! [14:31:13] and this one for all Unix pro's: http://www.flickr.com/photos/110698835@N04/11227971403/ ;) [14:33:22] Cyberpower678: btw. help is updated https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Configuring_MySQL_Workbench [14:33:39] hedonil: You should have linked to https://xkcd.com/1168/ instead. [14:34:05] anomie: ahh, the original one. [14:34:41] anomie: but it's a great cartoon - so full of truth ;) [14:38:16] anomie, :p [14:40:16] Cyberpower678: can you help me for a moment? [14:40:32] With what? [14:40:37] https://en.wikipedia.org/wiki/User:Xenocidic/statusChanger2.js [14:40:51] in here a line if (typeof(statusChangerConfig.statusList [14:41:24] my question is, Is this .statusList any fixed wikicode? Cyberpower678 [14:42:55] It's an array [14:43:09] Defined by the script [14:43:43] can I modify it? I mean can I take it to another language by changing it? Cyberpower678 [14:44:14] Sure. [14:45:03] I don't see why not. Unless I'm missing something. [14:45:11] Are having issues with it? [14:45:14] I'm having problem, Online's bengali is, অনলাইন . But when I use it the /Status produces %E0%A6%85%E0%A6%A8%E0%A62%E0%A6E%E0%A6%87%E0%A6%A8 [14:45:34] can you give me a solution? [14:46:07] Cyberpower678: [14:46:13] Umm.. I can't really work off of that. [14:46:19] Can I see the script? [14:47:49] yeah sure. https://test.wikipedia.org/wiki/User%3APratyya_Ghosh/Status-Chnger%28bn%29.js next time please ping me Cyberpower678 [14:49:56] anomie, just pinging you too. I'm not sure how to deal with this issue, since I've never encountered it. [14:50:19] anomie, being the genius you are. [14:51:56] Cyberpower678: "%E0%A6%85%E0%A6%A8%E0%A6%B2%E0%A6%BE%E0%A6%87%E0%A6%A8" is a percent-encoded representation of the UTF-8-encoded text "অনলাইন". So presumably something is encoding where it shouldn't or not decoding where it should. [14:52:13] Pratyya: Cyberpower678: Why not just use decodeURIComponent on output text ? [14:52:17] Pratyya, unicode characters seem to be getting urlencoded [14:53:15] Then what is the solution Cyberpower678 ? [14:54:34] hedonil suggested the fix in your case. The JS script is encoding the characters into a (URL format) how I like to call it. So it needs to be decoded again. [14:56:07] Pratyya, ^ [14:56:32] anomie: thank you for being the first in *ages* to formulate that correctly [14:56:49] You mean I need to decode it again. Okay... [14:57:21] So wrap outputs using the unicode characters in decodeURIComponent( URI ) [14:58:12] valhallasw: How so? By specifying that "অনলাইন" is UTF-8-encoded, instead of letting that encoding be assumed as the default per RFC 3986? [14:58:30] Pratyya, hey I can edit your js file. :p [14:58:39] * Cyberpower678 wonders why. [14:58:45] I know you are a sysop there. [14:58:53] * Cyberpower678 forgot [14:59:20] * Pratyya wonders what is happening to the user's here. just forget things. :P [14:59:42] anomie: Yes. And yes, since 2005 it should have been UTF-8, but there was a whole lot of internet before that ;-) [15:00:08] !log deployment-prep finished upgrading Elasticsearch in beta. We're on 0.90.9 now. [15:00:21] manybubbles: that was fast [15:00:41] Pratyya, you also have a logic error. [15:00:43] hashar: there aren't many indexes [15:00:44] manybubbles: can't wait for us to have a search.wikimedia.org that would use elastic as a backend [15:00:48] Logged the message, Master [15:00:58] ওহাত [15:01:08] oopps, what? Cyberpower678 [15:01:29] You changed stat to status which is undefined and might also be causing you issues. [15:01:31] hashar: doesn't that delegate to the search api on each wiki? [15:02:00] I think http://search.wikimedia.org is used for the open search api [15:02:08] got to pass some magic parameters to the url [15:02:18] would love to have a front page there that would be human friendly :D [15:03:03] actually I wanted to test things. I was wondering What happens if it was status. I can't do this at en wiki... They'll give a nice warning. :P [15:03:20] That's why doing all the breakings here :P Cyberpower678 [15:03:38] Pratyya, purge your cache and try again. [15:04:01] hashar: yeah. [15:04:14] btw, I just did the restarts for the beta elasticsearch machines. I'm off to do prod now. [15:05:27] no result is same. It brings %E0%A6%85%E0%A6%A8%E0%A62%E0%A6E%E0%A6%87%E0%A6%A8%E0%A7%87 [15:05:47] Ok. I haven't wrapped the output in decodeURIComponent [15:05:50] yet [15:05:51] you can test it by adding this to your common.js temporarily. [15:13:29] hey Cyberpower678 what is the difference between common.js and skin.js (skin=vector, monobook) ? [15:13:33] Pratyya, fixed [15:14:00] Pratyya, common.js takes effect regardless of what skin you have active and skin.js only takes affect if that skin is active. [15:14:34] Pratyya, purge your cache and test it now. It should work now. [15:14:43] ow, didn't know that. [15:15:16] Knowledge can be painful. ;p [15:15:48] ha ha ha =D, really sometimes it's "not" painful but painful to the power cube. :P [15:16:32] yeah it's fixed. Thank you. Cyberpower678 [15:16:57] petan: I have some questions about awstats [15:17:16] Pratyya, glad I could help. Be sure to thank hedonil and anomie [15:17:25] hedonil: ?? [15:17:37] hedonil: can you just ask them instead of asking if you can ask? :P [15:17:58] petan: some tools don#T show up in awstats. what is the issue with that? [15:18:12] Yeah sure. anomie and hedonil thanks to you guys. Without you this work could not be finished. Thanks. =D [15:18:21] Also Happy Happy 2014. [15:18:32] petan: is this related to newweb setting? [15:18:52] very likely yes [15:19:37] the new webservers that were configured by Coren do not write logs in the same format as apaches to the same location as they do [15:20:05] petan: would it be possible to put the /redacted/ logs from the proxy (all tools) into public share? [15:20:27] petan I would like to integrate them into my tool https://tools.wmflabs.org/wikiviewstats/ [15:20:53] Hey Cyberpower678 do you know that tool server is dead? [15:21:09] Pratyya, define dead. [15:22:15] The editcount isn't working at http://toolserver.org/~quentinv57/sulinfo/Pratyya_Ghosh [15:22:47] Works for me. [15:22:55] the wikidata edit count isn't updated. [15:22:57] See. [15:23:09] I have more than 200 edits, but it's 59 for me.. [15:23:24] Replication has been shut down on toolserver [15:23:45] I'm going to move that tool. [15:27:08] hedonil: yes that would be cool [15:27:19] hedonil: I will think of that [15:27:28] petan: great! [15:27:37] if you create a ticket maybe I will not forget immediately :D [15:27:50] petan: sure. will do. [15:29:04] Happy New Year again Cyberpower678 (I've posted a template in your talkpage) [15:29:18] Pratyya, you too [15:31:32] petan: or even better all labs webcounts, so everything can be put together and eyed in one place. I will add this to the ticket ;) [15:31:51] that is not easily doable [15:32:19] !log deployment-prep Migrated parsoid on deployment-parsoid2 to use mediawiki/services/parsoid out of a checkouts made in /srv/deployment/parsoid/{parsoid,deploy}. No job self updating it yet [15:32:45] Logged the message, Master [15:33:13] hey Cyberpower678 https://en.wikipedia.org/wiki/Template:Statustop in this page the offline used to beep as I remember. Why it doesn't beep now? [15:33:30] Beep? [15:33:42] I certainly hope not. [15:34:40] Beep means, how can I say this... ummmm... [15:34:45] jumping :p [15:34:57] * Pratyya wonders how can he say it. [15:35:08] * Cyberpower678 is lost [15:36:14] by the way how do you change your signatures? Is it any template? [15:36:17] Cyberpower678: [15:36:50] A very complex set of templates for the average Wiki editor. [15:37:20] * Cyberpower678 intends to convert that to a lua module for himself. [15:38:18] oh! I don't want any complex. Anyway do you come here @ IRC regularly Cyberpower678 [15:38:27] a ? is missing LP [15:38:41] Usually. [15:39:54] Bye today. Cyberpower678 . [15:40:00] Bye [15:40:09] You really helped me today, than you again for this. =D [16:43:05] Anybody want to help me work out why my xhr request to the wiki api is getting the help page? [16:43:36] what does your request look like? [16:44:53] Nettrom: Doesn't really matter, I can't seem to get it to work. Which makes me think there might be a problem in the JS instead. [16:45:17] does your request work OK if you send it manually to the API? [16:46:07] Nettrom: "Manually"? [16:46:29] a930913: yeah, type an example request into your browser's address bar and see if it works [16:46:54] Nettrom: It's the POST that's not working. [16:46:55] first stage of debugging, good way of checking of you've got the parameters right [16:48:32] Nettrom: xhr=new XMLHttpRequest();xhr.open("POST","//test.wikipedia.org/w/api.php?format=json",false);xhr.send("action=purge&titles=Main_Page"); [16:51:57] a930913: first: try add this line before send xhReq.setRequestHeader('Content-Type', 'application/x-www-form-urlencoded; charset=UTF-8'); [16:52:03] a930913: xhr=new XMLHttpRequest();xhr.open("POST","//test.wikipedia.org/w/api.php?action=purge&titles=Main_Page&format=json",false);xhr.send(); [16:52:33] not sure why you're POSTing that, though. [16:52:53] purge requires a POST request, should get an error back if you try GET [16:53:07] oh, needs POST if you're logged out [16:53:41] anyway, do a post but send the parameters as GET parameters [16:54:09] valhallasw: This is just a test. The real thing will need posts. [16:54:48] 'will need posts' [16:57:51] valhallasw: Hmm? [16:58:06] I'm not sure what you mean with that. [16:58:30] You mean this wouldn't work for page editing? [17:02:36] valhallasw: Yeah. [17:02:42] ok, after some fiddling, hedonil seems to be on the money [17:03:03] a930913: are your access control headers ok? [17:03:16] a930913: allow Cross-Origin Resource Sharing [17:03:27] xhr=new XMLHttpRequest();xhr.open("POST","//test.wikipedia.org/w/api.php",false);xhr.setRequestHeader("Content-Type", "application/x-www-form-urlencoded");xhr.send("action=purge&titles=Main_Page&format=json"); [17:03:33] ^ that works for me. [17:03:44] of course: *on* test.wikipedia.org [17:04:45] a930913: valhallasw: if you call test.wikipedia from domain wmflabs.org via AJAX you need additional headers [17:05:46] This will be a userscript, because the CORS stuff won't allow what I'm trying to do :( [17:06:10] valhallasw: a930913: if you are using toollabs new web there is an exmaple for .configuring your .lighttpd.conf [17:06:20] valhallasw: a930913: https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help/NewWeb#Examples_header.2C_mimetype.2C_error_handler [17:07:04] hedonil: that's if you want to call wikitech from enwiki, not the other way around [17:07:50] hedonil: I've spent near on a month trying to get CORS to play with webworkers one way or another. I've given up, and am just going to develop it directly on the wiki. [17:08:20] valhallasw: The content-type header seems to have done the trick. \o/ [17:08:59] * valhallasw points to hedonil, who came up with it originally ;-) [17:10:19] Heh, I must have glazed over that initially, thinking it not possible :p [17:10:29] valhallasw: yeah, fiddled with the same issues. these headers in .lighttpd.conf did the trick on CORS. but if you call it from same origin Content-Type setting is all you need ;) thx for the flowers ;) [17:11:44] I would have given up and used jquery :p [17:17:00] test [17:17:16] Clueless newbie here. Sorry if I get some of the terminology wrong. I'm on a Windows box. Got a Tool Labs shell account. Can connect to shell account using both PuTTY and WinSCP. I created a Tool. It shows up at tools.wmflabs.org. Would like to upload a html file to the public_html directory in that tool account. [17:17:27] The only way I can connect to my new tool is to connect to my user account with PuTTY, then "become TOOL". I cannot connect to the TOOL using WinSCP. How can I access public_html in the Tool account with WinSCP so I can easily upload a file to it? [17:22:14] padron: windows. with linux my suggestion would have been: sshfs [17:22:41] padron: you may want to try chaining as described here http://stackoverflow.com/questions/1010808/ssh-tunnelling-chain [17:24:19] padron: section: chaining with PuTTY, maybe it's similar to WinSCP - but to be honest, I've never tried it on my own. [17:24:42] padron: report, if you are successfull ;) [17:28:07] hedonil, I was sort of hoping I was making a simple mistake or missing something simple. What you point to seems like a lot of hoop jumping for what seemingly should be a simple thing. [17:28:57] padron: yeah, sad but true ;) [17:30:24] padron: if its a onetime job, create a public_html dir in user account, copy stuff, delete public_html from tools account, move public_html dir [17:33:13] padron: or even simpler copy to user account then cp -r /* /public_html/ [17:34:15] hedonil: Connecting and uploading files with WinSCP was a simple thing on toolserver. Are you saying things are different and not so simple in Labs? [17:35:33] padron: there's a saying: if it's documented, it's a feature not a bug ;) [17:36:37] padron: but of course you are right, this should be a simple routine [17:37:01] hedonil: Thank-you for your help. I will copy your suggestions and do my best. [17:37:46] padron: You should be able to upload to your tool's directories; in some cases (public_html) you need to "take FILE" (or "take DIR") afterwards as your tool to fix the ownership. [17:38:47] scfc_de: what's the trick to connect directly to tools directory? [17:39:19] scfc_de: in a simple way, and without chaining? [17:39:20] hedonil: Upload to /data/project/$TOOL/...? It should be writable from your user account. [17:39:34] scfc_de: The problem is that I cannot connect to or upload files to my tool's directories (with WinSCP). The only way I can access it is with PuTTY and "become TOOL". [17:40:23] scfc_de: as you see the problem is that your connection ends up at tools-login. [17:40:40] Yes, of course?! [17:41:45] But WinSCP ... there's a whole help page at https://wikitech.wikimedia.org/wiki/Help:Access_to_ToolLabs_instances_with_PuTTY_and_WinSCP. Is that useful for you, padron? [17:42:47] scfc_de: I read through that (and just about everything else). It helps me to connect to my user account.. but not the tool's account. [17:43:28] padron: Sure, but as your user account you should be able to write to your tool account's directories. [17:47:58] Hot dog I got it! From my user account I went UP in the directory structure into data, then project, then found my tool. Uploaded a file to public_html and all is well. [17:48:19] Never thought of doing that. [17:50:53] Thank-you hedonil and scfc_de for your assistance. I'm all squared away now and yes there is a simple way to do it ;-) [17:51:43] hedonil: BTW, regarding your AWStats question, there's a WIP at https://gerrit.wikimedia.org/r/#/c/80332/, but that depends on tools-webproxy being puppetized. IIRC this doesn't preserve the redacted log, but maybe the AWStats data is sufficient for you, or we could set up another pipeline. [17:52:58] padron: yes, was quite that easy, regards to scfc_de. [17:53:47] scfc_de: looking into this. I would /really/ like to integrate all these logs [17:56:41] scfc_de: it there an ETA? [17:58:36] scfc_de: the current awstats statistics are lacking of some newweb configured tools and bots [18:05:11] hedonil: I hope to have finished puppetizing tools-webproxy by the end of the week (the AWStats stuff is done IIRC). As -webproxy is the heart of the web access, review & deployment will then be *very* careful :-), but in general, depending on Coren's schedule, it should be deployable by between the end of next week and the end of the month. [18:06:07] scfc_de: that sounds like good news [18:27:49] I keep getting an error message: MediaWiki internal error. [18:27:50] Exception caught inside exception handler. [18:27:50] Set $wgShowExceptionDetails = true; at the bottom of LocalSettings.php to show detailed debugging information. [18:27:57] on commons [18:30:19] janstee: Me, too, and the folks over at #wikimedia-operations seem to be quite busy, so maybe there's something going on ... :-) [18:30:53] thanks for the news - [20:02:43] hedonil, I've got a question. What's a decent replacement for scanning the toolserver's wiki to obtain all of the projects?\ [20:03:35] Cyberpower678: example? [20:04:33] An old tool on toolserver would use an SQL query like "SELECT `dbname`,`server`,`domain` FROM `wiki` WHERE `is_closed`=0 ORDER BY `server`" [20:05:11] Cyberpower678: ah, it's database meta_p on every slice [20:05:58] Cyberpower678: now you can check with MySQL Workbench ! ;) [20:06:16] I know. I'm setting up the connection now. :D [20:12:02] hedonil, meta_p or metawiki_p? [20:12:18] Cyberpower678: meta_p [20:12:25] Host? [20:12:35] Cyberpower678: any host [20:12:44] Okay [20:14:10] hedonil, Error Code: 1548 Cannot load from mysql.proc. The table is probably corrupted [20:14:30] Cyberpower678: huh [20:15:32] That's what happened when I tried to access the meta_p database. [20:15:42] Cyberpower678: which host? [20:15:55] metawiki.labsdb [20:16:08] * hedonil tries [20:16:30] hedonil, I tried numerous times. [20:17:16]