[00:01:34] <hedonil>	 ireas: this is really a good idea. many could work on that and many could profit from that.
[00:01:46] <ireas>	 hedonil, yes!
[00:02:36] <ireas>	 Krinkle, ah, one question that came to my mind some weeks ago when looking Intuition: what’s the best way to create a new message for my tool in translatewiki? (I never used translatewiki before…)
[00:06:08] <Krinkle>	 ireas: Messages are created in git, not on translatewiki.net
[00:06:26] <Krinkle>	 translatewiki.net will sync with the source code, offer them to translators and sync back to source code
[00:07:06] <ireas>	 Krinkle, ah, okay, that explains why I couldn’t find the option ‘add translation’ xD
[00:07:08] <ireas>	 thanks!
[00:08:09] <Krinkle>	 You can register a new tool andd/or add/change messages via https://github.com/Krinkle/TsIntuition
[00:08:15] <Krinkle>	 fork, commit, pull request
[00:08:25] <Krinkle>	 Siebrand and I will review and merge asap
[00:12:12] <Krinkle>	 ireas: And to test it locally before submitting, you can include the tool from the central location, and load your draft tool-i18n file like this: https://github.com/Krinkle/TsIntuition/wiki/Documentation#register-entire-text-domain-locally 
[00:12:44] <ireas>	 Krinkle, okay, thanks. I’ll have to check what I implemented already and what is still missing ^^
[00:13:24] <Krinkle>	 Then once it works and the pull request is approved, you'd remove that temporary load line, and instead load let intuition load it from the central install path on that server
[00:17:24] <hedonil>	 Krinkle: That's it! Intuition - love at second sight.
[00:35:39] <Coren>	 Sorry about vanishing without warning:  my dog decided to rip out a claw and I had to rush him to the vet.
[00:53:22] <cajoel>	 gerrit project creation request: operations/puppet/netflow 
[01:40:42] <Betacommand>	 legoktm: you around?
[01:41:04] <legoktm>	 hi
[01:42:42] <Betacommand>	 legoktm: same thing wiht Apocatequil
[01:42:45] <legoktm>	 lol
[01:42:46] <legoktm>	 ok
[01:44:39] <legoktm>	 done
[01:49:29] <Betacommand>	 thanks
[01:55:16] <Scottywong>	 howdy, anyone home?  trying to get ssh access to tools lab, but running into some issues, I get:
[01:55:21] <Scottywong>	 Permission denied (publickey)
[01:55:29] <Scottywong>	 not sure where I'm going wrong
[01:57:15] <Coren>	 Scottywong: Things to check:
[01:57:37] <Coren>	 (a) shell user names are case significant, and all-lowercase.
[01:57:59] <Scottywong>	 yup, I used all lowercase
[01:58:03] <Coren>	 (b) that your key is either pointed to by your client, or in your agent.
[01:59:23] <Coren>	 Scottywong: I'm not seeing any attempt from you in the logs.  What host are you connecting to?
[01:59:33] <Scottywong>	 hmm I assume it is, I'm using the Terminal in mac os x, just using the ssh command, and my key is stored in $HOME/.ssh
[01:59:43] <Scottywong>	 connecting to tools-login.wmflabs.org
[02:00:26] <Coren>	 Can you try now?
[02:00:36] <Coren>	 Whilst I'm watching.
[02:02:05] <Scottywong>	 just did it
[02:02:26] <Coren>	 How odd, I'm not seeing your attempt at all.  Wait, what username are you trying?
[02:02:32] <Scottywong>	 scottywong
[02:02:46] <Scottywong>	 ssh scottywong@tools-login.wmflabs.org
[02:03:08] <Scottywong>	 hmm when I uploaded my SSH key to my preferences, should I include the "ssh-rsa" at the beginning of it?  or just the actual key part
[02:03:24] <Coren>	 You need the whole thing, including the key type.
[02:03:58] <Coren>	 But I'm not seeing you failing your attempt, I'm not seeing you /try/.  Can you do 'ssh -vv scottywong@tools-login.wmflabs.org' and pastebin the result?
[02:11:27] <Scottywong>	 looks like I might be having trouble accessing the key file locally for some reason
[02:11:31] <Scottywong>	 perhaps
[02:12:18] <Scottywong>	 oh crap, that might be my fault, I think I moved it to a different location.  lemme try moving it to the right directory and trying again
[02:12:31] <Coren>	 Please try 'ssh -p 222 scottywong@tools-login.wmflabs.org' and pastebin the result?
[02:12:49] <Coren>	 Ah.
[02:12:50] <Coren>	 :-)
[02:14:03] <Scottywong>	 that worked
[02:14:07] <Scottywong>	 sorry for being dumb :)
[02:15:07] <Coren>	 Erare humanum est.
[02:33:03] <Betacommand>	 legoktm: Ive got another if you have a second
[02:43:22] <jeremyb>	 ryan_lane is the 4th google hit for "debian landscape" :-]
[02:44:37] <legoktm>	 Betacommand: sure
[02:44:51] * jeremyb  wonders what this mysterious task is :-P

[02:45:08] <Betacommand>	 http://en.wikipedia.org/wiki/Wikipedia:Bots/Status/inactive_bots_1 eagle 101 and 10 need updated
[02:58:01] <ireas>	 is there any way to access the apache error.log for a tool?
[03:00:00] <legoktm>	 ireas: see https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Web_logs tl;dr: no
[03:00:23] <ireas>	 legoktm, ok, thanks.
[03:06:55] <Scottywong>	 another noob question:  I created a new tool, and then I enter "become toolname" and I get this response:
[03:07:03] <Scottywong>	 sudo: sorry, a password is required to run sudo
[03:08:21] <legoktm>	 Scottywong: need to log out and login again
[03:10:08] <Scottywong>	 oh ok i'll try that
[03:10:34] <Scottywong>	 well, that was easy
[03:10:35] <Scottywong>	 thanks
[03:29:03] <LuaKT>	 I feel dumb, I'm running a query on the enwiki database, if I copy and paste a string from one webpage to match I get an empty set, if I copy the same string from another page I get results...That's not possible right?
[03:30:49] <legoktm>	 umm
[03:30:53] <legoktm>	 what's the query you're running?
[03:33:29] <LuaKT>	 Oh if I paste it here shows a space after the string, guess it was a dodgy character I couldn't see on cygwin..
[03:43:28] <SigmaWP>	 Scottywong: Hi
[03:43:41] <Scottywong>	 howdy
[05:33:21] <mutante>	 needs a way to edit data in a mysql database as a group, based on labs project membership
[05:33:37] <mutante>	 and no, no phpmyadmin
[05:34:19] <mutante>	 s/mysql/mariadb
[05:35:30] <mutante>	 heh, i guess what i actually want is people to put queries into gerrit changes and they're executed when merged
[10:23:40] <hashar>	 !log deployment-prep rebooting bunch of servers
[10:25:39] <hashar>	 !log deployment-prep restarting udp2log-mw , bunch of defunct python demuxer.py 
[10:25:43] <hashar>	 bah poor bot
[10:25:44] <hashar>	 ping
[10:25:47] <hashar>	 wm-bot: ping
[10:25:47] <wm-bot>	 Hi hashar, there is some error, I am a stupid bot and I am not intelligent enough to hold a conversation with you :-)
[10:48:05] <hedonil>	 !newweb
[10:48:05] <wm-bot>	 https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help/NewWeb
[11:10:47] <GerardM->	 Hoi, the replication of Wikidata (probably Wikipedia-de as well) is broken 
[11:13:45] <Betacommand>	 GerardM-: check the mailing list
[15:14:49] <hashar>	 !log deployment-prep made /home/jenkins-deploy readable and accessible by group and others: chmod +rx /home/jenkins-deploy
[15:48:50] <andrewbogott>	 mhoover: How are things going?  Are the next steps forward clear, and are you blocked by anything?
[16:08:52] <mhoover>	 andrewbogott: hey man! i'm cool on my vm machine - not blocked. installing stack parts on some test machines i have, and soon to install on my vm. next steps would be testing puppet n friends with stack. almost there
[16:09:30] <mhoover>	 andrewbogott: and when that's all cool, start on the hardware in va
[16:09:51] <andrewbogott>	 mhoover: ok, great.  At some point you'll need reviews to get your new puppet stuff merged in -- I should be available all week for that.
[16:10:32] <mhoover>	 andrewbogott: sweet. i can stumble through most of it, but will most likely need your assistance :)
[16:10:54] <andrewbogott>	 well, by 'review' I mostly meant +2
[16:11:22] <mhoover>	 andrewbogott: hehe
[16:11:46] <mhoover>	 andrewbogott: fyi, i'll be traveling most of the day tomorrow
[16:11:52] <andrewbogott>	 ok
[16:11:59] <mhoover>	 andrewbogott: but the rest of the week is good
[16:14:05] <mhoover>	 andrewbogott: btw, what dhcp space do you typically allocate for the labs vm's? /16? /20?
[16:14:16] <mhoover>	 andrewbogott: more?
[16:15:13] <andrewbogott>	 hm, I don't actually know.
[16:15:23] <andrewbogott>	 Although, you can log on to virt2 and check :)
[16:16:31] <mhoover>	 andrewbogott: roger :)
[16:18:15] <hashar>	 !log deployment-prep refreshed Parsoid source code in /srv/deployment/parsoid , deleted node_modules and reinstalled them
[16:18:17] <hashar>	 bah
[16:18:18] <hashar>	 morebots
[16:18:20] <hashar>	 r:(
[16:18:58] <mhoover>	 andrewbogott: is there a network/service map somewhere of the openstack config you currently have or want to have?
[16:20:21] <andrewbogott>	 mhoover: I doubt it!  Ryan would know.
[16:20:27] <andrewbogott>	 hashar:  I'll give morebots a kick
[16:21:05] <andrewbogott>	 mhoover: I came in in the middle so would need to reverse-engineer the design same as you :(
[16:21:25] <andrewbogott>	 (which I'm happy to do if it helps)
[16:22:18] <andrewbogott>	 labs-morebots, feeling better?
[16:22:18] <labs-morebots>	 I am a logbot running on tools-exec-02.
[16:22:18] <labs-morebots>	 Messages are logged to wikitech.wikimedia.org/wiki/Server_Admin_Log.
[16:22:18] <labs-morebots>	 To log a message, type !log <msg>.
[16:51:03] <andrewbogott>	 back soon...
[16:58:29] <hashar>	 zeljkof: it is getting worse, parsoid gives me a 404 :D
[16:58:30] <hashar>	 Error loading data from server: parsoidserver-http-bad-status: 404. Would you like to retry?
[18:55:31] <anomie>	 Coren: Hmm. This is entirely not urgent, but why did job 1743953 not exit already? It should have run for only about one second.
[18:57:29] <Coren>	 40030    19861  0.0  0.0  17560  1476 ?        SNs  01:48   0:00  |   \_ /usr/bin/perl -we sleep 1
[18:59:35] <Coren>	 anomie: Quoting.  the command you passed to perl is 'sleep' without arguments, followed by '1' into @ARGV.
[18:59:45] <Coren>	 "sleep: Causes the script to sleep for (integer) EXPR seconds, or forever if no argument is given."
[18:59:48] <anomie>	 Coren: Yeah, I just realized that. d'oh.
[19:00:18] <Coren>	 In related news, "WTF Larry?  When is sleep forever useful?"
[19:00:33] <Coren>	 Or at least useful enough to be the default behaviour.  :-)
[19:00:48] <anomie>	 Coren: I tried to qdel it and qdel told me it couldn't resolve "tools-master.pmtpa.wmflabs". Then I tried again and it worked.
[19:01:00] <Coren>	 o_O odd.
[19:01:24] <Coren>	 Hm.  tools-login is ridiculously loaded.
[19:01:32] <Coren>	 Someone's running bots on it again.
[19:03:37] <valhallasw>	 kill --with-fire
[19:10:09] * anomie  sees pid 1969 on tools-login and thinks "WTF?"

[19:18:14] <ToAruShiroiNeko>	 hi
[19:18:22] <ToAruShiroiNeko>	 is anyone up for a query on en.wikipedia?
[19:18:32] <ToAruShiroiNeko>	 I want a process of indef blocked IPs
[19:18:39] <Coren>	 anomie: Meh, someone doing work on a dump; not all that unusual.
[19:19:53] <Coren>	 ToAruShiroiNeko: I'm a bit busy to help with running a query atm; but I'm sure someone can help you here.  Barring that, getting access yourself is fast and easy.  :-)
[19:20:44] <ToAruShiroiNeko>	 I lack to time to really focus on it
[19:21:03] <ToAruShiroiNeko>	 I want to have a list of indef blocked ips without edits for example
[19:21:20] <ToAruShiroiNeko>	 or perhaps less than 50 edits
[19:22:09] <Coren>	 That's actually harder than it seems, certainly not as simple as a query (because of rangeblocks), and it won't catch IPs or ranges with non-anonymous contribs (like socks)
[19:22:30] <ToAruShiroiNeko>	 there are 20k single IP blocks
[19:22:40] <ToAruShiroiNeko>	 those are what I want to work on
[19:22:49] <ToAruShiroiNeko>	 https://meta.wikimedia.org/wiki/User:%E3%81%A8%E3%81%82%E3%82%8B%E7%99%BD%E3%81%84%E7%8C%AB/English_Wikipedia_open_proxy_candidates
[19:22:55] <ToAruShiroiNeko>	 you can process that list if you like
[19:23:01] <ToAruShiroiNeko>	 or output to it
[19:26:33] <Coren>	 ToAruShiroiNeko: Sure, but not all will have been blocked because of anon edits.   https://en.wikipedia.org/wiki/Wikipedia:Bot_requests is probably the best place to ask, though.
[19:28:15] * valhallasw  hits gerrit with a shovel

[19:35:40] <ToAruShiroiNeko>	 Coren its not a bot request
[19:35:50] <ToAruShiroiNeko>	 I just need output for sysops to see
[19:36:00] <Matthew_>	 valhallasw: Be gentle.  Gerrit is sketchy enough as it is XD
[19:36:42] <Coren>	 I know, but generating reports of the sort is the kind of thing bot people do; especially since you'll probably want to have that report run again in the future, and it's not likely to be answered with just a plain SQL query.
[20:11:28] <ToAruShiroiNeko>	 Coren oh no
[20:11:30] <ToAruShiroiNeko>	 its a one time thing
[20:11:37] <ToAruShiroiNeko>	 we wont get 20k single ip blocks ever again
[21:04:48] <brainwane>	 hi ksnider, hope things are going well.
[21:05:24] <ksnider>	 Hi Sumana! 
[21:05:26] <ksnider>	 How's things?
[21:09:07] <brainwane>	 Pretty great, ksnider... I am currently investigating the joys of LAMP-type apps
[21:10:06] <petan>	 andrewbogott_afk, Coren: I can't ssh to huggle
[21:10:13] <petan>	 what's wrong :/
[21:10:54] <petan>	 but apache is running there
[21:11:16] <petan>	 mutante?
[21:11:31] <petan>	 @notify Ryan_Lane
[21:11:31] <wm-bot>	 I'll let you know when I see Ryan_Lane around here
[21:12:15] <andrewbogott>	 petan, is 'huggle' the instance name or project name or both?
[21:12:21] <petan>	 both
[21:12:27] <andrewbogott>	 ok, I'll have a look...
[21:12:38] <petan>	 I just can't ssh
[21:14:52] <petan>	 at some point it's not so critical, I needed to fix cluebot's HAN line, but cluebot is still down
[21:14:59] <andrewbogott>	 petan, try again?
[21:15:21] <petan>	 HA
[21:15:22] <petan>	 now it works
[21:16:24] <andrewbogott>	 great!
[21:17:05] <mutante>	 eh, hi
[21:17:18] <mutante>	 already fixed is good to read
[21:31:59] <brainwane>	 I'm trying to get my Flask web app to actually display at https://tools.wmflabs.org/missing-from-wikipedia/datainput.html . I presume I'm supposed to put the Flask application (because it's Python) into public_html -- anyone wanna take a look and help?
[21:32:11] <brainwane>	 (Coren?)
[21:32:50] <Coren>	 brainwane: In theory, yes.  What roadblock are you hitting?
[21:33:33] <brainwane>	 Coren: so, https://tools.wmflabs.org/missing-from-wikipedia/datainput.html gives me a nice 404
[21:34:06] <Coren>	 brainwane: /missing-from-wikipedia/ does map to your tool's public_html
[21:34:09] <brainwane>	 yeah
[21:34:12] <brainwane>	 so that works
[21:34:21] <brainwane>	 I put a basic index.html page there
[21:34:53] <brainwane>	 I could move around my templates folder so both my .html files are directly in public_html
[21:35:15] <brainwane>	 I'll try that
[21:35:27] <ireas>	 brainwane, you should not access the templates directly
[21:35:32] <Coren>	 Well, you can keep it in a subdir like you do now but then the URL becomes (predictably) https://tools.wmflabs.org/missing-from-wikipedia/missing-from-wikipedia/datainput.html which is not so helpful.  :-)
[21:35:57] <ireas>	 brainwane, or did you define a 'datainput.html' route?
[21:36:04] <valhallasw>	 !newweb
[21:36:04] <wm-bot>	 https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help/NewWeb
[21:36:25] <Coren>	 brainwane: Also ireas and valhallasw both have more python-fu than I do.
[21:36:27] <Coren>	 :-)
[21:36:36] <Coren>	 Listen to them in preference to me.
[21:36:47] <valhallasw>	 brainwane: I have used fcgi for the gerrit patch uploader (which is also a flask app)
[21:37:02] <valhallasw>	 but I just saw something was missing on the NewWeb page on how to do that
[21:37:16] <brainwane>	 ireas: I did not define a datainput.html route - https://git.wikimedia.org/blob/mediawiki%2Ftools%2Fmissing-from-wikipedia/54675e9f44421d16c86cee8f6734ae2dc4c7e705/webapp%2Fapplication.py
[21:37:22] <brainwane>	 the route is for index
[21:37:49] <ireas>	 brainwane, so you want to open /missing-from-wikipedia/index, don’t you?
[21:37:54] <brainwane>	 hmmmmmmmmm
[21:38:41] <brainwane>	 all my variants fail
[21:38:47] <brainwane>	 e.g. https://tools.wmflabs.org/missing-from-wikipedia/missing-from-wikipedia/webapp/index
[21:39:04] <ireas>	 valhallasw, ah, looks great! is the fastcgi server started automatically?
[21:39:13] <valhallasw>	 ireas: yes, by lighttpd
[21:39:16] <brainwane>	 I could temporarily remove the current index.html file, lemme try that (https://tools.wmflabs.org/missing-from-wikipedia/index)
[21:39:23] <ireas>	 :-)
[21:39:37] <ireas>	 brainwane, how did you setup your flask application? CGI? FCGI?
[21:40:01] <brainwane>	 ireas: I just stuck my whole git repo, including the Flask app, into the public_html directory
[21:41:04] <ireas>	 brainwane, oh, okay. then you have to configure how to execute the flask application -- either via CGI as described here http://flask.pocoo.org/docs/deploying/cgi/, or using FCGI as described on the newweb page linked by valhallasw
[21:41:11] * ireas  tries FCGI now

[21:41:38] <valhallasw>	 ireas: https://github.com/valhallasw/gerrit-patch-uploader/blob/master/app.fcgi
[21:42:38] <brainwane>	 ok, I did webservice start
[21:42:45] <brainwane>	 now https://tools.wmflabs.org/missing-from-wikipedia/missingfromwikipedia/index gives me a DIFFERENT 404, which is good
[21:42:59] <brainwane>	 same for https://tools.wmflabs.org/missing-from-wikipedia/index
[21:43:28] <valhallasw>	 brainwane, copy https://github.com/valhallasw/gerrit-patch-uploader/blob/master/index.py to your public_html
[21:43:41] <valhallasw>	 brainwane: and adapt the from app import app so it imports your app
[21:43:53] <valhallasw>	 if you chmod 755 it, you should be able to access it from the web
[21:44:20] <valhallasw>	 (it's probably easiest to put it in /data/project/missing-from-wikipedia/public_html/missing-from-wikipedia/webapp , and then to do 'from application import app')
[21:44:34] <valhallasw>	 then the app should be available under http://tools.wmflabs.org/missing-from-wikipedia/missing-from-wikipedia/webapp/index.py/index
[21:45:23] <valhallasw>	 see also:   ls /data/project/gerrit-reviewer-bot/public_html/patchuploader/
[21:46:00] <brainwane>	 valhallasw: when you say "easiest to put it in", do you mean it's easiest to put that index.py file into that dir?
[21:46:25] <brainwane>	 (that's what I infer)
[21:46:29] <valhallasw>	 Correct
[21:47:50] <valhallasw>	 but it doesn't seem to work :/
[21:48:25] <brainwane>	 valhallasw: I just restarted my webservice
[21:50:08] <brainwane>	 ok! interesting! ireas & valhallasw - do I actually need to set the application running manually with, say, 'python application.py' or does the webservice take care of that? currently I see that instead of the HTML template rendering, the raw HTML is displaying in the browser, and I wonder whether it's related to the fact that the webservice is running AND I've manually got 'application.py' running
[21:50:48] <ireas>	 valhallasw, just accessing the index.py that calls CGIHandler will not work, you have to configure Apache too (at least in my case)
[21:51:04] <valhallasw>	 brainwane: index.py will be run by the web server, and will in turn load your app from application.py
[21:51:26] <valhallasw>	 ireas: you are right!
[21:51:54] <valhallasw>	 brainwane: you'll also have to cp /data/project/gerrit-reviewer-bot/public_html/patchuploader/.htaccess
[21:51:57] <brainwane>	 http://tools.wmflabs.org/missing-from-wikipedia/missing-from-wikipedia/webapp/index.py/index  is at least not a 404! but is not as interesting as it ought to be. OK, ireas, do I also need to.... ok
[21:52:00] <valhallasw>	 and adapt that to the correct URL
[21:52:21] <valhallasw>	 fcgi was actually easier to get to work... :/
[21:52:49] <ireas>	 brainwane, doesnt look so bad! but it seems you have a manual `print "we did a get"` statement somewhere. remove that
[21:52:55] <ireas>	 valhallasw, wohoo, works! :) https://tools.wmflabs.org/pb/pb-beta/
[21:53:33] <valhallasw>	 ireas: that's with fcgi?
[21:53:37] <ireas>	 valhallasw, yes
[21:53:39] <valhallasw>	 cool :-)
[21:54:19] <brainwane>	 ireas: some debugging stuff; now commented out.
[21:54:38] <ireas>	 brainwane, so now it works! :)
[21:54:52] <brainwane>	 ok! thank you valhallasw and ireas ! I guessed at the .htaccess stuff and it mostly worked! ... now an error 500.
[21:54:59] * brainwane  turns Debug=True on again

[21:55:28] <brainwane>	 if this is a Unicode problem again I am gonna hit something
[21:55:50] <valhallasw>	 brainwane: have you read http://www.joelonsoftware.com/articles/Unicode.html ? :-)
[21:55:55] <brainwane>	 I have indeed 
[21:56:03] <brainwane>	 :)
[21:56:45] <valhallasw>	 but Flask should do most of the heavy lifting -- afaik all in/output is in unicode strings
[21:57:57] <brainwane>	 hm, I turned Debug to True and it is still giving me a 500 Internal Server Error
[21:58:32] * brainwane  tries to look at ~/error.log

[21:59:42] <brainwane>	 nothing useful. in fact it doesn't have any errors at all
[22:00:10] <brainwane>	 https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help/NewWeb#Enabling_request_logging - do I add this line in ~/.lighttpd.conf ?
[22:01:08] <giftpflanze>	 is flask just an obscure python thingy that you don't need (and cannot use) with other languages?
[22:01:25] <brainwane>	 Flask is a micro webapp framework for Python, giftpflanze 
[22:01:28] <brainwane>	 similar to Sinatra in Ruby
[22:02:22] <Scottywong>	 is tools.wmflabs.org down?  seems to be taking forever to load simple html files, and times out about 80% of the time.  not sure if it's just me...
[22:02:40] <brainwane>	 I'm having that problem too Scottywong but I thought it was just my application!
[22:03:28] <ireas>	 giftpflanze, flask is Python only, but *very* cool :-)
[22:03:44] <giftpflanze>	 i still don't get it :)
[22:04:03] <brainwane>	 giftpflanze: you may not realize that you came off as kind of dismissive, btw
[22:04:03] <giftpflanze>	 why would you need that?
[22:04:25] <brainwane>	 giftpflanze: the usual use case is: quickly get a little CRUD-type webapp up and running
[22:04:48] <ireas>	 giftpflanze, because there are some tasks you don’t want to do from scratch for every single web application
[22:05:03] <giftpflanze>	 i might be ignorant, yes
[22:05:06] <ireas>	 (cookie handling, request routing, logging, template loading ...)
[22:05:34] <brainwane>	 so, ireas, now that the general tool labs slowdown has kind of passed, I am stymied by the 500 error. Any tips? there's nothing in ~/.error_log, I turned on Debug=True in my main function and yet am still just getting a plain error 500...
[22:06:38] <valhallasw>	 brainwane: unfortunately, I can't help you much more as I'm going to bed. If you need more assistance, please let me know tomorrow :-)
[22:10:46] <brainwane>	 So, ok. my current hypothesis is that FCGI is eating my stack trace or something. Coren, where should I be looking for the real error logs? I look in ~/.error_log and just see a few lines about application restarting from like 30 min ago, nothing recent and none of the 500s
[22:15:29] <Scottywong>	 is there an ETA on when the tools web server will be reliably operational?  i can't seem to even reliably load tiny html files without timing out most of the time.
[22:15:57] <Scottywong>	 or even load the main tools.wmflabs.org home page
[22:16:07] <Scottywong>	 without timing out
[22:17:40] <ireas>	 Scottywong, no answer to your question, but HTTPS works better than HTTP
[22:18:28] <Scottywong>	 oh?  i'll give that a try
[22:19:03] <Scottywong>	 ahh, that does seem to work somewhat better, thanks
[22:19:36] <brainwane>	 yeah
[22:20:02] <brainwane>	 also Scottywong I know I heard Coren say that a change coming soon will make things faster in general, can't remember what the change will be
[22:34:21] <hashar>	 !log bastion ping
[22:34:28] <labs-morebots>	 Logged the message, Master
[22:34:47] <hashar>	 !log deployment-prep rerunning puppet continuously on deployment-cache-mobile01 + apt-get upgrade of varnish
[22:34:50] <labs-morebots>	 Logged the message, Master
[22:35:26] <Scottywong>	 anyone have experience running python scripts as CGI on tools lab?  I keep getting an internal server error.  Tried it in ~/public_html and ~/cgi_bin, no difference.  It has execute bits set, and it is owned by the tool account.
[22:35:45] <Scottywong>	 I tried running it from the shell, and it runs just fine without errors.
[22:36:07] <hashar>	 !log deployment-prep Parsoid got broken since last Friday (serving pages from production …) {{bug|57926}}
[22:36:11] <labs-morebots>	 Logged the message, Master
[22:36:25] <hashar>	 !log deployment-prep rebooting deployment-cache-mobile01 
[22:36:30] <labs-morebots>	 Logged the message, Master
[22:37:15] <Scottywong>	 even tried renaming it to .py and still got an internal server error
[22:41:45] <hashar>	 Scottywong: no clue
[22:41:53] <hashar>	 Scottywong: but you can ask on the labs-l mailing list
[22:42:16] <hashar>	 will probably not give you an answer tonight, but might get one tomorrow :d
[22:42:36] <hashar>	 !log deployment-prep upgrading packages on deployment-cache-bits03 && rebooting
[22:42:41] <labs-morebots>	 Logged the message, Master
[22:43:00] <Scottywong>	 i imagine there aren't any interesting error logs I could look at, are there?
[22:43:09] <Scottywong>	 i looked in ~/access.log but it's not useful
[22:43:14] <hashar>	 if you have access to logs, it is worth a look at 
[22:43:37] <hashar>	 i don't know how cgi works nor howtoolslabs is setup sorry
[22:47:34] <Scottywong>	 no worries
[23:06:26] <Betacommand>	 Elsie: are your DBR scrips in an svn repo?
[23:10:21] <Coren>	 Scottywong: Still around?
[23:17:08] <Coren>	 Scottywong: If you use the new web service setup ...
[23:17:11] <Coren>	 !newweb
[23:17:11] <wm-bot>	 https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help/NewWeb
[23:17:28] <Coren>	 You get (a) much improved performance and (b) error logs delivered to your $HOME
[23:17:29] <Coren>	 :-)
[23:22:54] <brainwane>	 I'm having trouble "take"ing a file 
[23:23:05] <brainwane>	 $ take /home/brainwane/biographies.txt
[23:23:05] <brainwane>	 /home/brainwane/biographies.txt: you must own the containing directory
[23:23:19] <brainwane>	 where do I put this file so my tools acct can "take" it?
[23:23:23] <brainwane>	 petan: any ideas?
[23:25:28] <andrewbogott>	 tools accounts usually have their own home directory
[23:25:32] <Coren>	 brainwane: In any directory your tool account owns.
[23:25:45] <andrewbogott>	 in, I think… /data/project/<toolname>
[23:25:45] <brainwane>	 ok, so I should have scp
[23:25:46] <Coren>	 I.e. mv it before you take it.  :-)
[23:25:53] <brainwane>	 I tried....
[23:26:07] <Coren>	 Ah, your own home doesn't give permission to your tool.  :-)
[23:26:09] <brainwane>	 $ mv biographies.txt /data/project/missing-from-wikipedia/public_html/missing-from-wikipedia/
[23:26:09] <brainwane>	 mv: cannot create regular file `/data/project/missing-from-wikipedia/public_html/missing-from-wikipedia/biographies.txt': Permission denied
[23:26:25] <Coren>	 brainwane: Ah!  Your permissions are too strict then.
[23:26:29] <brainwane>	 oh?
[23:26:53] <Coren>	 brainwane: You probably want your tool's home to be g+w anyways; otherwise you can't properly have more than one maintainer without a lot of trouble.
[23:27:10] <brainwane>	 g+w? groupwrite?
[23:27:11] <Coren>	 brainwane: from your tool account, "chmod -R g+w ~"
[23:27:39] <brainwane>	 got it. done
[23:27:44] <Coren>	 That'll give your tool maintainers write access to the tool, and that includes yourself.  :-)
[23:27:58] <brainwane>	 cool. was that in the docs and I missed it?
[23:28:22] <Coren>	 That's normal unixy stuff which we mostly don't cover in the labs-specific docs.
[23:28:39] <brainwane>	 :/
[23:28:42] <Coren>	 It would be good if we could find a nice tutorial we could link to though.
[23:28:50] <brainwane>	 I've been using Unix since 1999 and I missed this.
[23:29:08] <brainwane>	 I presumed that creating a tool account would automatically give me as maintainer the necessary permissions.
[23:29:40] <Coren>	 brainwane: It does; but if you did things like use some scp clients or unpacked tarballs with restrictive permissions, it'll override.
[23:30:36] <brainwane>	 ok. I added a thing to the help page's talk page in hopes someone will do an update to the docs at some point.
[23:30:45] <brainwane>	 gotta head off. thanks
[23:36:11] <ireas>	 someone online who’s familiar with newweb / lighttpd? I need the lighttpd version of Apache’s DirectoryIndex for a single directory
[23:37:07] <mutante>	 !lighttpd
[23:37:08] <wm-bot>	 http://redmine.lighttpd.net/projects/lighttpd/wiki/Tutorials
[23:38:30] <mutante>	 ireas: this looks like it  http://redmine.lighttpd.net/projects/1/wiki/Docs_ModDirlisting
[23:38:54] <mutante>	      dir-listing.activate = "enable" 
[23:41:17] <ireas>	 mutante, no, it’s about changing the file that is accessed by default (default: '/' → '/index.html', what I need: '/' → 'index.py')
[23:42:19] <Coren>	 ireas: You could use a rewrite rule.
[23:42:33] <ireas>	 Coren, you are right, I’ll try that
[23:42:35] <ireas>	 thanks!
[23:42:37] <mutante>	 ireas: index-file.names = ( "index.html" )
[23:43:02] <mutante>	 sorry, i mixed that up with Options Indexes
[23:43:13] <Coren>	 mutante has the actually correct solution, too.
[23:43:46] <ireas>	 mutante, no problem & thanks!
[23:45:17] <ireas>	 hmmm, that somehow breaks my fastcgi configuration ...
[23:46:55] <ireas>	 tools.wmflabs.org/paste/view/10dd0def without the first line, it works, with the first line, lighttpd fails and stops
[23:47:05] <ireas>	 2013-12-03 23:44:30: (configfile.c.912) source: /var/run/lighttpd/pb.conf line: 560 pos: 15 parser failed somehow near here: (EOL)
[23:49:53] <hedonil>	 ireas: did you try += instead of = ?
[23:50:26] <ireas>	 hedonil, in the fastcgi line? no, bot if I omit the index-file line, everything works … I’ll try it though
[23:50:56] <hedonil>	 ireas: in the index-file line
[23:51:32] <ireas>	 hedonil, ah, I mistaked it. one moment …
[23:52:30] <ireas>	 hedonil, wow, it helped! thanks! do you know *why* = did not work?
[23:54:12] <hedonil>	 ireas: the merge constantly fails with only '=' if the parameter is already defined in master config. I guess.
[23:54:48] <ireas>	 hedonil, hm, sounds reasonable. I’ll add this as a notice to the /NewWeb page.