[00:12:16] <jeremyb>	 Brooke: i think maybe that was a reference to https://en.wikipedia.org/wiki/User:Jasper_Deng/IPv6
[00:12:39] <Brooke>	 I don't think that was quoted in her post.
[00:12:45] <Brooke>	 But that's the one I was thinking of.
[00:12:50] <Brooke>	 I thought it was on Meta-Wiki.
[00:12:56] <Brooke>	 But at Meta-Wiki he properly used the main namespace.
[00:12:58] <Brooke>	 So.. 
[00:13:06] <Brooke>	 I guess a wikitech-l feed in here might be nice.
[00:13:09] <Brooke>	 Mayne.
[00:13:11] <Brooke>	 b
[00:15:25] <jeremyb>	 Maine?
[00:16:06] <jeremyb>	 and a wikitech-ambassadors and a /me tries to think of something else
[01:35:31] <MJ94>	 User talk page modification: error "Gateway Time-out" occurred while contacting the API.
[01:43:25] <BarkingFish>	 Evening guys, are you aware of any issues with Wikimedia's API? People are reporting errors in tools which use it
[01:43:42] <BarkingFish>	 Gateway time out, warning messages from the API and so on
[01:44:49] <Romaine>	 here too
[01:45:31] <BarkingFish>	 I would check for probs myself, but I'm using a text only browser, and the status page and nagios don't show up properly in it.
[01:45:48] <BarkingFish>	 I'm minus a GUI at the moment, which slows shit down dramatically for me :(
[01:46:11] <SigmaWP>	 https://en.wikipedia.org/w/api.php?action=query&list=recentchanges&rctoken=patrol&rclimit=1&format=jsonfm
[01:46:22] <SigmaWP>	 Says I don't have the 'patrol' right, and I'm logged in
[01:46:48] <MJ94>	 BarkingFish: done :)
[01:47:48] <BarkingFish>	 MJ94: it's fixed?
[01:48:03] <MJ94>	 Nope, I just meant done, I alreay allerted them
[01:48:06] <MJ94>	 :)
[01:48:29] <BarkingFish>	 Ah, ok then - sorry for bugging you dudes, I'll leave you in peace. I'll keep watching for any notes though :)
[01:53:13] <MJ94>	 BarkingFish: Haha, no one was listening to me, so you helped.
[01:53:56] <BarkingFish>	 MJ94: they listen, believe me.  The reason nobody responds is that most likely, they're up to their necks in trying to work out what went feet up :)
[01:54:05] <BarkingFish>	 I've learned that from a long time reporting stuff in here :)
[01:54:12] <MJ94>	 hahaha yes
[01:54:43] <MJ94>	 BarkingFish: I consider myself pretty good with computers, but I'm no problem solvers, so hi5 to the techies.
[02:07:13] <anomie>	 Is anyone else seeing 504 Gateway Time-out from api.php? Request: GET http://en.wikipedia.org/w/api.php, from 10.64.0.131 via cp1005.eqiad.wmnet (squid/2.7.STABLE9) to ()
[02:07:13] <anomie>	 Error: ERR_CANNOT_FORWARD, errno (11) Resource temporarily unavailable at Fri, 08 Jun 2012 02:06:08 GMT 
[02:07:52] <SigmaWP>	 anomie: Yes
[02:07:55] <BarkingFish>	 anomie: yep, errors on the API have been reported here :)
[02:08:00] <BarkingFish>	 and elsewhere too :P
[02:08:06] <SigmaWP>	 VPT, for instance
[02:08:18] <SigmaWP>	 And everyone who had the patience to hear me complain
[02:14:55] <logmsgbot_>	 !log LocalisationUpdate completed (1.20wmf4) at Fri Jun  8 02:14:55 UTC 2012
[02:15:02] <morebots>	 Logged the message, Master
[02:15:10] <SigmaWP>	 !log Hello, world!
[02:15:16] <morebots>	 Logged the message, Master
[02:15:20] <SigmaWP>	 ._.
[02:15:23] <AndrewN>	 o.O
[02:15:26] <Bsadowski1>	 Don't do that, SigmaWP.
[02:15:27] <SigmaWP>	 Er
[02:15:28] <AndrewN>	 O.O
[02:15:30] <SigmaWP>	 Sorry!
[02:15:32] <Bsadowski1>	 :|
[02:15:37] <SigmaWP>	 Everyone type /clear please
[02:15:55] <AndrewN>	 A wild Bsadowski1 appears!
[02:23:09] <logmsgbot_>	 !log LocalisationUpdate completed (1.20wmf3) at Fri Jun  8 02:23:09 UTC 2012
[02:23:13] <morebots>	 Logged the message, Master
[02:43:18] <logmsgbot_>	 tstarling cleared profiling data
[02:50:46] <TimStarling>	 is the only problem that the API is down?
[02:51:06] <jeremyb>	 seemed more widespread from nagios
[02:51:21] <jeremyb>	 i think all the complaints have been about the API
[02:51:37] <TimStarling>	 oh charming
[02:51:41] <Betacommand>	 http://en.wikipedia.org/w/api.php
[02:51:46] <TimStarling>	 gmond is using 100% CPU and 23GB RSS
[02:51:52] <Betacommand>	 Our servers are currently experiencing a technical problem.
[02:52:56] <jeremyb>	 SigmaWP: fyi, that dents and tweets and maybe does other stuff too. i reverted the one place i could for you. http://wikitech.wikimedia.org/index.php?title=Server_admin_log&diff=47798&oldid=47797
[02:53:37] <TimStarling>	 !log on cp1002: killed gmond, which was using 100% CPU and 23GB RSS. Restarting squid which had died
[02:53:42] <morebots>	 Logged the message, Master
[02:56:21] <TimStarling>	 !log cp1001: same as on cp1002, restarted gmond
[02:56:25] <morebots>	 Logged the message, Master
[02:56:39] <jeremyb>	 but they've been doing that for a good 4 days at least. why now? was it just building the whole time?
[02:56:56] <jeremyb>	 (could be 10 days even for all i know)
[02:57:03] <SigmaWP>	 jeremyb: That dents tweets? :O
[02:57:11] <jeremyb>	 SigmaWP: dents and tweets
[02:57:20] <SigmaWP>	 Oh, ok
[02:57:24] <TimStarling>	 ganglia does not show any useful data for that cluster so it's hard to tell
[02:57:26] * SigmaWP  WHAT HAVE I DONE?

[02:57:31] <SigmaWP>	 jeremyb: Thanks for the revert
[02:57:49] <jeremyb>	 SigmaWP: no big deal. just thought you should know it's not just a wiki. there's real people following the streams
[02:58:38] <jeremyb>	 yeah, it's got a small blip of data
[02:58:45] <SigmaWP>	 OK
[02:59:15] * jeremyb  waves Ryan_Lane

[02:59:22] <jeremyb>	 api issues
[02:59:35] <jeremyb>	 tim booted gmond on both cp100[12]
[02:59:41] <jeremyb>	 oh, i guess that's not really a Ryan_Lane
[02:59:50] <jeremyb>	 he's automated
[03:00:33] <jeremyb>	 (i was wondering, it's kinda early there! if he's still there)
[03:01:23] <TimStarling>	 !log on fenari: copied *.text and *.upload from /home/wikipedia/conf/squid/generated/clusters to /etc/dsh/group
[03:01:28] <morebots>	 Logged the message, Master
[03:02:54] <TimStarling>	 !log on cp1002: killed gmond again, it was leaking memory again, already up to 27GB in the few minutes since I restarted it
[03:02:58] <morebots>	 Logged the message, Master
[03:08:19] <jeremyb>	 TimStarling: so what's happening?
[03:08:27] <TimStarling>	 !log stopped gmond on cp1001 with kill -STOP for memory leak debugging
[03:08:32] <morebots>	 Logged the message, Master
[03:08:34] <jeremyb>	 heh
[03:08:35] <TimStarling>	 is the API working again now?
[03:08:51] <String[]>	 Let's try it.
[03:08:57] <jeremyb>	 String[]: Betacommand Bsadowski1 anomie ping
[03:08:59] <anomie>	 No, still down.
[03:09:05] <Betacommand>	 jeremyb: pong
[03:09:09] <jeremyb>	 ^^^
[03:09:12] <jeremyb>	 try the API
[03:09:35] <String[]>	 TimStarling: No
[03:09:47] <Betacommand>	 works about 50% of the time
[03:10:25] <anomie>	 Interestingly enough, if I hack /etc/hosts to locally point en.wikipedia.org to 208.80.152.201 (wikipedia-lb.pmtpa.wikimedia.org) instead of 208.80.154.225 (wikipedia-lb.eqiad.wikimedia.org), it seems to work ok.
[03:10:25] <Betacommand>	 TimStarling: take a look at cp1005 also
[03:10:59] <jeremyb>	 anomie: what is it otherwise?
[03:11:21] <anomie>	 jeremyb- I get eqiad here, 208.80.154.225
[03:11:22] <String[]>	 Error 504: Gateway Time-out, if anyone cares
[03:11:27] <jeremyb>	 Betacommand: you're getting an error there? same one anomie reported before
[03:11:33] <jeremyb>	 String[]: right
[03:12:12] <Betacommand>	 jeremyb: Im getting the wiki has a problem page and rror: ERR_CANNOT_FORWARD, errno (11) Resource temporarily unavailable at Fri, 08 Jun 2012 03:10:10 GMT 
[03:12:51] <jeremyb>	 Betacommand: yeah, but why did you say cp1005? is that what the error said?
[03:12:58] <Betacommand>	 via cp1005.eqiad.wmnet
[03:13:15] <anomie>	 The error always mentions cp1005.eqiad.wmnet, as far as I've seen.
[03:13:25] <String[]>	 Hrm...
[03:13:36] <anomie>	 I don't know enough about things to know if that's unusual
[03:13:40] <jeremyb>	 it's amazing how unuseful http://ganglia.wikimedia.org/latest/?c=Text%20squids%20eqiad&h=cp1005.eqiad.wmnet&m=load_one&r=hour&s=by%20name&hc=4&mc=2 is ;-P
[03:13:44] <String[]>	 I just changed the &format to jsonfm, and it worked
[03:13:55] <String[]>	 Nevermind, just a 1-time thing
[03:14:03] <jeremyb>	 String[]: example URL?
[03:14:10] <String[]>	 https://en.wikipedia.org/w/api.php?action=query&prop=templates&tllimit=500&tlnamespace=10&titles=Wikipedia:Sandbox&format=jsonfm
[03:14:18] <jeremyb>	 danke
[03:14:43] <Betacommand>	 jeremyb: http://en.wikipedia.org/w/api.php fails for me
[03:14:51] <jeremyb>	 i'm gettign a lot of cp1004 here
[03:14:54] <jeremyb>	 getting*
[03:15:20] <Bsadowski1>	 "Request: GET http://en.wikipedia.org/w/api.php, from 10.64.0.127 via cp1005.eqiad.wmnet (squid/2.7.STABLE9) to ()
[03:15:20] <Bsadowski1>	 Error: ERR_CANNOT_FORWARD, errno [No Error] at Fri, 08 Jun 2012 03:15:02 GMT "
[03:15:23] <Bsadowski1>	 :O
[03:15:31] <jeremyb>	 yup
[03:16:05] <TimStarling>	 maybe being out of memory broke it
[03:16:22] <jeremyb>	 but these boxes aren't running gmond
[03:16:30] <jeremyb>	 (cp100[45])
[03:17:57] <Betacommand>	 Im seeing errors from 004 about half of the refreshes
[03:18:05] <TimStarling>	 !log restarting squid on cp1005, maybe out of FDs or something, cachemgr shows exactly 1000 open connections to 10.2.1.1
[03:18:09] <morebots>	 Logged the message, Master
[03:18:41] <jeremyb>	 what's that # on cp1004?
[03:18:52] <kgb[]>	 hello everyone
[03:19:10] <jeremyb>	 kgb[]: what's your error today? ;)
[03:20:50] <kgb[]>	 I would like to create gallery site based on *all* wikimedia images. What is better: to crawl entire wikicommons with bot and download them OR just hardlinkthem from my site to wikicommon servers? What its better from the perspective of wikicommons server [im aware that i might be causing big load or "stealing bandwitch"]. thats why im asking before trying. ps im aware of the size of wikicommons
[03:21:21] <jeremyb>	 well you definitely shouldn't just crawl
[03:21:28] <jeremyb>	 if you want them all you can rsync them
[03:21:57] <jeremyb>	 also make it wikimedia commons or just "commons" but not wikicommons
[03:22:05] <anomie>	 Starting to see errors mentioning cp1003.eqiad.wmnet now.
[03:22:49] <kgb[]>	 ok. didint knwo that u got rsync :)
[03:22:55] <kgb[]>	 sorry for typos
[03:23:09] <kgb[]>	 so running rsync is ok yeah? and i will not risk ban?
[03:23:15] <jeremyb>	 kgb[]: you should join xmldatadumps-l
[03:23:25] <jeremyb>	 no, rsync is fine
[03:23:35] <jeremyb>	 @search lists
[03:23:35] <wm-bot>	 Results (Found 7): 1.19, announce, list, lists, repeat, smw, support, 
[03:23:40] <jeremyb>	 !list
[03:23:40] <wm-bot>	 mediawiki-l and wikitech-l are the primary mailing lists for MediaWiki-related issues. See http://lists.wikimedia.org/ for details.
[03:23:40] <snitch>	 meta.wikimedia; page; Tech
[03:23:43] <jeremyb>	 gah
[03:23:47] <jeremyb>	 !lists
[03:23:47] <wm-bot>	 mediawiki-l and wikitech-l are the primary mailing lists for MediaWiki-related issues. See https://www.mediawiki.org/wiki/Mailing_lists for details.
[03:24:13] <jeremyb>	 !mailarchive is http://lists.wikimedia.org/pipermail/$1
[03:24:13] <wm-bot>	 Key was added
[03:24:56] <jeremyb>	 !listinfo is https://lists.wikimedia.org/mailman/listinfo/$1
[03:24:57] <wm-bot>	 Key was added
[03:25:04] <jeremyb>	 !pipermail alias mailarchive
[03:25:04] <wm-bot>	 Created new alias for this key
[03:25:19] <jeremyb>	 !listinfo xmldatadumps-l | kgb[]
[03:25:19] <wm-bot>	 kgb[]: https://lists.wikimedia.org/mailman/listinfo/xmldatadumps-l
[03:25:41] <kgb[]>	 thx buddy
[03:25:44] <kgb[]>	 let me browse it
[03:27:23] <jeremyb>	 kgb[]: http://bots.wmflabs.org/~petrb/logs/%23wikimedia-tech/20120606.txt starting at 22:01:07
[03:29:10] <kgb[]>	 nice :)
[03:30:34] <anomie>	 Assuming the X-Cache and X-Cache-Lookup also correspond to whichever cp100x.eqiad.wmnet host is involved, cp1005 is still giving me failures (and seems to be hit most often). Infrequently cp1005 gives me a successful response. cp1003 works most of the time, but sometimes not. Also have seen cp1004, cp1002, and cp1001, no errors from them so far for me.
[03:31:02] <jeremyb>	 cp1004 is ging me plenty of errors
[03:31:17] <TimStarling>	 if you don't want to use cp1005, then don't post to http://en.wikipedia.org/w/api.php, use some other URL
[03:31:31] <jeremyb>	 giving*
[03:33:18] <kgb[]>	 jeremyb: http://dumps.wikimedia.org/commonswiki/20120601/ Which one contains the images? From what i see those are only dumps with metadata?
[03:33:47] <anomie>	 What else is giving problems besides the API?
[03:33:47] <jeremyb>	 did you read the log?
[03:33:59] <kgb[]>	 more or less
[03:34:00] <jeremyb>	 http://ftpmirror.your.org/pub/wikimedia/images/wikipedia/commons/
[03:34:02] <jeremyb>	 gah
[03:34:03] <Betacommand>	 kgb[]: there isnt a image dump
[03:34:24] <kgb[]>	 ok then
[03:35:15] <kgb[]>	 "As of November 2011 the image and other media files take up about 17T, most of it already compressed media." :)
[03:35:25] <jeremyb>	 hah
[03:36:27] <kgb[]>	 is there any ls -lR file avalible from ftp mirror?
[03:37:58] * jeremyb  is digging a little...

[03:38:57] <TimStarling>	 it's clicktracking
[03:39:09] <jeremyb>	 heh
[03:39:17] <jeremyb>	 a new deploy?
[03:39:27] <Brooke>	 meta.wikimedia.org seems painfully slow to me.
[03:39:31] <jeremyb>	 makes sense that it would be API
[03:40:19] <jeremyb>	 kgb[]: so, that http link above corresponds to rsync://ftpmirror.your.org/wikimedia-images/wikipedia/commons/ ; you can browse http://ftpmirror.your.org/pub/wikimedia/images/ to see which other parts of the corresponding rsync tree you want to fetch
[03:42:09] <jeremyb>	 kgb[]: what site is this for?
[03:46:48] <jeremyb>	 ah, the e3 thing was a recent deploy and that uses clicktracking
[03:47:50] <TimStarling>	 what e3 thing?
[03:48:25] <TimStarling>	 LastModified/E3Experiments/js/ext.E3Experiments.Timestamp.js ?
[03:48:29] <jeremyb>	 yes
[03:48:35] <jeremyb>	 it was on for only testwiki from may 31 and then kaldari turned it on today for enwiki
[03:49:04] <TimStarling>	 well, it's going away now
[03:49:21] <jeremyb>	 i think you can just turn it out in initialisesettings
[03:49:27] <jeremyb>	 and that should do it
[03:49:47] <jeremyb>	 !g I292efae75418
[03:49:47] <wm-bot>	 https://gerrit.wikimedia.org/r/#q,I292efae75418,n,z
[03:50:07] <jeremyb>	 just back that out
[03:51:03] <logmsgbot_>	 !log tstarling synchronized wmf-config/InitialiseSettings.php 
[03:51:07] <morebots>	 Logged the message, Master
[03:53:49] <kgb[]>	 jeremyb: i wanna create some "gallery" website and i want to steal wikicommons content lol
[03:54:16] <jeremyb>	 kgb[]: <jeremyb> also make it wikimedia commons or just "commons" but not wikicommons
[03:54:56] <jeremyb>	 kgb[]: do you understand all of the legal ramifications? i.e. how to properly credit image authors on your site?
[03:55:07] <TimStarling>	 !log disabled LastModified extension due to overload on cp1005
[03:55:12] <morebots>	 Logged the message, Master
[03:55:47] <Eloquence>	 did the e3 team take the site down already?
[03:55:48] <jeremyb>	 TimStarling: did you push to gerrit yet?
[03:55:48] <kgb[]>	 jeremyb: well im not an expert but yeah i understand it more or less.
[03:55:53] <jeremyb>	 Eloquence: yup ;)
[03:55:56] <TimStarling>	 I'm working on it
[03:56:25] <jeremyb>	 Eloquence: mostly API. but at least 5 people noticed it and complained here. surely they weren't the only ones to notice
[03:56:27] <TimStarling>	 the change is in the local repo on fenari, I'm working on pushing it to gerrit
[03:56:48] <TimStarling>	 Eloquence: every page view was leading to a POST to the API
[03:57:13] <Eloquence>	 gah
[03:57:26] <jeremyb>	 TimStarling: only one per view?
[03:57:27] <TimStarling>	 squid broke first, but if I fixed squid, apache probably would have broken instead
[03:58:59] <TimStarling>	 it's hard to imagine the apache cluster being able to support such a high request rate
[03:59:49] <jeremyb>	 i wonder how it lasted so long? it was deployed ~8 hrs ago
[03:59:50] <Brooke>	 Any idea if it was the JS or the PHP hitting the API?
[04:00:21] <jeremyb>	 err, 7.5 hrs rather
[04:00:28] <jeremyb>	 Brooke: JS?
[04:00:33] <jeremyb>	 why would php use the api?
[04:00:54] <Brooke>	 You can do FauxRequests, I think.
[04:00:57] <jeremyb>	 i know there's those faux requests. but those don't leave the box even, right?
[04:01:10] <jeremyb>	 that's why they're called faux
[04:01:10] <Brooke>	 Not sure why the JS would be posting all the time.
[04:01:23] <Brooke>	 They still hit the API, I think.
[04:01:34] <jeremyb>	 for clicktracking. to report datapoints to be recorded
[04:01:42] <jeremyb>	 what does e3 mean anyway?
[04:01:52] <Brooke>	 It seems like it d be better to disable the ClickTracking extension.
[04:01:58] <Brooke>	 Editor engagement experiments
[04:02:03] <Brooke>	 it'd
[04:02:41] <jeremyb>	 ahh
[04:03:09] <TimStarling>	 Eloquence: I'll post an incident report to a mailing list
[04:03:24] <TimStarling>	 maybe engineering
[04:04:00] <Bidgee|Away>	 Unserstand that you maybe busy ATM, I'm wondering how I request a server-side upload of some large vids?
[04:04:11] <Eloquence>	 TimStarling, thanks.
[04:04:15] <jeremyb>	 Bidgee|Away: file a bug
[04:04:37] <Brooke>	 https://bugzilla.wikimedia.org
[04:04:56] <jeremyb>	 Bidgee|Away: provide an http(s) URL where the files can be downloaded from. provide a .txt file per file with the description page contents
[04:05:13] <jeremyb>	 Bidgee|Away: file 1 single bug for the batch of files once that's all ready
[04:05:21] <Brooke>	 https://meta.wikimedia.org/wiki/Tech
[04:05:27] <Brooke>	 Two threads at the bottom of the page if anyone wants them.
[04:06:08] <Bidgee|Away>	 Cheers, I've uploaded them onto Archive.org
[04:06:18] <jeremyb>	 great
[04:06:23] <jeremyb>	 now you have to make description pages
[04:07:07] <snitch>	 [[Tech]]; MZMcBride; /* Help regarding an analysis tool! */ +reply; https://meta.wikimedia.org/w/index.php?diff=3817730&oldid=3814452&rcid=3337341
[04:07:23] <Brooke>	 > Few Indic language wikipedians have asked a feature (for the respective language wiki) in Special:Upload page to make sure that the users are selecting a license and type a description before uploading images. Do we have a feature in mediawiki to support this? I have seen similar feature in Commons where some of the actions are mandatory. Is this is a Gadget/extension/something else?--Shiju Alex (WMF) (talk) 11:27, 5 June 2012 (UTC)
[04:07:29] <Brooke>	 Anyone know that one?
[04:07:43] <Brooke>	 en.wiki made its own file upload wizard. Commons has UploadWizard.
[04:08:25] <Eloquence>	 Brooke, en.wiki's wizard is sauron while commons' is gandalf.
[04:08:44] <Brooke>	 Heh, I'm not even nerd enough to understand that.
[04:08:48] <Bidgee|Away>	 Cheers again. Thank you for your help! :)
[04:09:47] <Brooke>	 Sauron --> bad; Gandalf --> good.
[04:12:58] <jeremyb>	 Bidgee|Away: https://bugzilla.wikimedia.org/enter_bug.cgi?bug_severity=normal&bug_status=NEW&component=Site%20configuration&keywords=shell&product=Wikimedia&short_desc=Batch%20upload%20large%20files%3A%20XYZ%20lorem%20ipsum
[04:13:25] <jeremyb>	 Bidgee|Away: how big are they btw?
[04:14:16] <anomie>	 Just in case confirmation is needed, API seems to be working now here. Goodnight!
[04:15:08] <Bidgee|Away>	 http://archive.org/details/OpalsPressConferenceAtAisWithLaurenJacksonCarrieGrafAndJennaOhea Part 4 is the only one that is below the 100mb limit
[04:15:28] <jeremyb>	 Bidgee|Away: k. file away
[04:15:35] <k>	 no
[04:15:45] <jeremyb>	 Bidgee|Away: (let me know when you're done)
[04:15:47] <jeremyb>	 k: ?
[04:16:07] <k>	 ?!
[04:16:08] <jeremyb>	 Bidgee|Away: err, wait
[04:16:55] <jeremyb>	 Bidgee|Away: you could upload 2 yourself. the limit's now 500 if you turn on the experiment
[04:17:04] <jeremyb>	 Bidgee|Away: but 2 you'd still need to file for
[04:17:06] <jeremyb>	 for now
[04:17:26] <Bidgee|Away>	 Done that but still said 100mb :(
[04:17:47] <jeremyb>	 where?
[04:17:51] <jeremyb>	 what did you do?
[04:18:49] <jeremyb>	 https://commons.wikimedia.org/wiki/Commons:Chunked_uploads
[04:19:58] <Bidgee|Away>	 Turned on the experiment in the UploadWizard Preferences 
[04:21:32] <Bidgee|Away>	 hrm
[04:22:24] <Bidgee|Away>	 It worked this time. No idea why it didn't work before
[04:23:07] <jeremyb>	 cool
[04:23:19] <jeremyb>	 anyway, pls point me to the bug for the 2 big once
[04:23:21] <jeremyb>	 ones*
[04:23:32] <Bidgee|Away>	 http://archive.org/download/OpalsPressConferenceAtAisWithLaurenJacksonCarrieGrafAndJennaOhea/MVI_5423.ogv needs to be uploaded over http://commons.wikimedia.org/wiki/File:Opals_press_conference_at_AIS_with_Lauren_Jackson,_Carrie_Graf_and_Jenna_O%27Hea_%28part_1%29.ogv
[04:24:15] <jeremyb>	 Bidgee|Away: what is it, lower res?
[04:24:50] <Bidgee|Away>	 yer, so it could fit under the 100mb limit
[04:25:11] <jeremyb>	 also, why not make it all a single video?
[04:25:15] <jeremyb>	 not 4 parts?
[04:25:35] <Bidgee|Away>	 http://archive.org/download/OpalsPressConferenceAtAisWithLaurenJacksonCarrieGrafAndJennaOhea/MVI_5424.ogv http://commons.wikimedia.org/wiki/File:Opals_press_conference_at_AIS_with_Lauren_Jackson,_Carrie_Graf_and_Jenna_O%27Hea_%28part_2%29.ogv
[04:26:47] <Bidgee|Away>	 My computer will not allow me to do it as it is four different videos (DSLR only allows ~10min recording time)
[04:27:20] <jeremyb>	 oh, ewww
[04:27:42] <Bidgee|Away>	 I've got the software to due it but not the hardware (being HD it really eats the memory and HDD space up)
[04:27:58] <jeremyb>	 yeah
[04:29:49] <Bidgee|Away>	 Hoping that there is a hack that will override the limit in the future as it seems Canon isn't going to remove it. 
[04:30:29] <jeremyb>	 what do you mean, it's not just you ran out of disk?
[04:30:31] <jeremyb>	 what camera?
[04:32:30] <jeremyb>	 Bidgee|Away: http://publiclaboratory.org/wiki/camera-trigger
[04:34:00] <Bidgee|Away>	 Canon EOS 60D
[04:34:36] <Bidgee|Away>	 It has a limitation in the camera's firmware. 
[04:35:48] <Bidgee|Away>	 It has its own "Record" button
[04:37:50] <jeremyb>	 yeah, just you made me think of it
[04:38:55] <Bidgee|Away>	 Looking at using http://magiclantern.wikia.com/
[04:39:41] <jeremyb>	 how about t3i? ;)
[04:39:46] <jeremyb>	 does it have that problem?
[04:40:40] <Bidgee|Away>	 Yep
[04:42:34] <Bidgee|Away>	 Even the Nikon has the same limit
[04:49:53] <jeremyb>	 crazy
[04:50:07] <jeremyb>	 maybe ok for shooting a movie. but not for live events!
[05:31:17] <StevenW>	 TimStarling: do we have any idea how LastModified killed the API? (As in any particular component.)
[05:34:33] <Brooke>	 StevenW: I'm told it was related to ClickTracking.
[05:35:06] <StevenW>	 Sounds likely. 
[05:36:06] <Brooke>	 https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/LastModified.git;a=blob;f=E3Experiments/js/ext.E3Experiments.Timestamp.js;h=78f0731f4a3c15131163ce78b352c525313c1a23;hb=1519f4e7e1b3cde1cd3045deae6e50bf8c73ccb9
[05:36:15] <Brooke>	 Maybe that?
[05:37:07] <StevenW>	 Yes, because IIRC it was not sampling impressions but logging them all. 
[05:37:30] <StevenW>	 In the past, all of that has been sampled at a low rate. 
[05:37:42] <Brooke>	 Who reviewed the code?
[05:38:25] <StevenW>	 Roan reviewed the bucketing and clicktracking part, Kaldari reviewed LastModified proper.
[05:38:45] <Brooke>	 Roan didn't notice the API hit? Hm.
[05:51:28] <TimStarling>	 StevenW: the post data was action=clicktracking&format=json&eventid=ext.lastModified%401-ctrl1-impression&token=...&namespacenumber=0
[06:50:19] <TimStarling>	 !log on cp1001: disabled HTCP plugin in gmond for testing, seems to work so I will disable it properly
[06:50:24] <morebots>	 Logged the message, Master
[11:29:16] <Akoopal>	 mark: around?
[11:29:36] <Akoopal>	 (affraid not as you were talking about weekend yesterday)
[14:01:40] <vvv>	 Any op around?
[14:01:47] <vvv>	 We have Russian Planet broken
[14:18:58] <sumanah>	 https://bugzilla.wikimedia.org/show_bug.cgi?id=37408 -- a DejaVu thumbnail isn't being created at upload.wikimedia.org, EOF error
[14:19:03] <sdv_>	 here i am
[14:19:53] * sumanah  waits for operations folks to take a look

[14:29:00] <sumanah>	 notpeter: ping
[14:29:18] <sumanah>	 (hope you are enjoying your travel!)
[14:33:43] <ToAruShiroiNeko>	 DejaVu?
[14:35:14] <sumanah>	 yeah
[14:35:40] <sdv_>	 Yes. Moreover, in the bug report I didn't write that, if you lower 5100px to e.g. 3000px, the image appears
[14:36:32] <sumanah>	 sdv_: add it?
[14:36:43] <sumanah>	 sdv_: you can just enter that in the "comment" area
[14:36:49] <sdv_>	 I don't know how to edit the message
[14:37:02] <sdv_>	 ah ok i'll add a new comment
[14:37:19] <sumanah>	 yeah
[14:37:52] <sumanah>	 if you look at other bug reports (example: https://bugzilla.wikimedia.org/show_bug.cgi?id=35046 ) you see that people keep adding information with new comments, and follow up with questions.
[14:48:33] <sumanah>	 apergos: https://bugzilla.wikimedia.org/show_bug.cgi?id=37408 -- a DejaVu thumbnail isn't being created at upload.wikimedia.org, EOF error .  Should I wait till Aaron Schulz comes online?
[14:58:38] <apergos>	 yes and I don't know how much he knows about djvu format thumbs either but I surely know even less :-D
[14:59:01] <sumanah>	 damn.
[14:59:11] <sumanah>	 I was wondering whether this was related to the Swift changeover
[15:33:55] <sumanah>	 sdv: so, I need to go
[15:34:20] <sumanah>	 sdv: this might be related to how ProofreadPage (the extension) processes or displays the thumbnails? but I don't know
[15:34:36] <sumanah>	 other people here and in #mediawiki and on Bugzilla will know more.
[15:50:54] <snitch>	 [[Tech]]; MarcoAurelio; /* Checkboxes at Special:UserRights — Order changed for no reason? */ new section; https://meta.wikimedia.org/w/index.php?diff=3818754&oldid=3817730&rcid=3337930
[17:34:27] <Alchimista>	 domas: around?
[17:35:19] <domas>	 maybe
[17:35:36] <domas>	 not for too long
[17:37:44] <Alchimista>	 domas: would it be difficult for you to make avaiable the Page view info of a single project?
[17:40:13] <domas>	 yes
[17:40:15] <domas>	 maybe
[17:41:50] <Alchimista>	 it would be quite usefull, i need to track the development of a small project, with aprox 700 pages, so it would be a nightmare to download the big datafiles, just to get some few data
[18:14:05] <snitch>	 [[Tech]]; MZMcBride; /* Checkboxes at Special:UserRights — Order changed for no reason? */ +reply; https://meta.wikimedia.org/w/index.php?diff=3818980&oldid=3818754&rcid=3338025
[18:47:25] <logmsgbot_>	 !log reedy synchronized php-1.20wmf4/languages/messages/  'Pushing out updated files upon siebrands request'
[18:47:29] <morebots>	 Logged the message, Master
[18:47:36] <Reedy>	 There's a scap needed for that
[18:47:40] <Reedy>	 will do it in a little bit...
[19:39:30] <logmsgbot_>	 !log reedy Started syncing Wikimedia installation... : Rebuilding localisation cache for message updates
[19:39:34] <morebots>	 Logged the message, Master
[19:40:22] <Reedy>	 Updating ExtensionMessages-1.20wmf3.php...
[19:40:22] <Reedy>	 Database name bnwikimedia is not listed in pmtpa.dblist
[19:40:22] <Reedy>	 Updating LocalisationCache for 1.20wmf3... Database name bnwikimedia is not listed in pmtpa.dblist
[19:40:22] <Reedy>	 done
[19:40:38] <omtsh>	 O_O
[19:57:28] <logmsgbot_>	 !log reedy Finished syncing Wikimedia installation... : Rebuilding localisation cache for message updates
[19:57:33] <morebots>	 Logged the message, Master
[19:58:00] <jeremyb>	 20 minutes?!
[20:00:44] <Reedy>	 Yeah, the concurrency was massively reduced
[20:01:54] <jeremyb>	 still...
[20:02:01] <jeremyb>	 does the nightly one take so long?
[20:02:16] <jeremyb>	 bbl
[20:04:55] <logmsgbot_>	 !log reedy rebuilt wikiversions.cdb and synchronized wikiversions files: 2 wikimedia wikis to 1.20wmf4
[20:05:00] <morebots>	 Logged the message, Master
[20:07:55] <Reedy>	 Bah, sync-dblist hasn't been merged yet
[20:08:25] <logmsgbot_>	 !log reedy synchronized all.dblist 
[20:08:29] <morebots>	 Logged the message, Master
[20:08:51] <jeremyb>	 Reedy: btw, idk who it was but i see you syncing now... last night dblists were out of sync with gerrit
[20:09:06] <Reedy>	 Whar do you mean?
[20:09:13] <logmsgbot_>	 !log reedy synchronized wikimedia.dblist 
[20:09:17] <morebots>	 Logged the message, Master
[20:09:17] <Reedy>	 I just noticed chwikimedia and bnwikimedia weren't in all or wikimedia
[20:09:20] <jeremyb>	 someone changed wikiversions without commiting it
[20:10:15] <Reedy>	 I haven't touched it before today since... monday?
[20:11:02] <jeremyb>	 prolly aaron on the 6th at 16 UTC http://wikitech.wikimedia.org/index.php?title=Server_admin_log&diff=47781&oldid=47780
[20:11:24] <Reedy>	 heh
[20:11:47] <jeremyb>	 git-deploy will fix that i hope
[20:13:56] <Reedy>	 bleh
[20:18:23] <Reedy>	 deb rebuilds, pffttt
[20:32:35] <Reedy>	 !log Updated php to point to php-1.20wmf4 rather than php-1.20wmf3
[20:32:39] <morebots>	 Logged the message, Master
[21:29:40] <Brooke>	 https access to Wikimedia wikis seems slow and high latency in the past few days.
[21:30:03] <Reedy>	 seems ok in europeland
[21:30:05] <Brooke>	 Shit just seems to be taking longer than it used to.
[21:30:12] <Brooke>	 When I click a talk page tab or just navigate around.
[21:30:14] <Brooke>	 It'll just sit there.
[21:30:21] <Brooke>	 Might be my connection.
[21:33:46] <Betacommand>	 Reedy: do you have a second to run a database query for me, I want to see if the toolserver is having data corruption or if its a mediawki issue
[21:34:05] <Reedy>	 Possibly
[21:34:07] <Betacommand>	 select count(*) from templatelinks where tl_namespace = 10 and tl_title = "Don't_know";
[21:34:25] <Reedy>	 enwiki?
[21:34:28] <Betacommand>	 yeah
[21:34:40] <Reedy>	 255
[21:34:56] <Betacommand>	 ok, then we have a mediawiki issue
[21:35:10] * Betacommand  goes to file a bug report