[00:31:19] gn8 folks [00:37:42] [[Tech]]; MiszaBot; Robot: Archiving 2 threads (older than 30d) to [[Tech/Archives/2013]].; https://meta.wikimedia.org/w/index.php?diff=5483069&oldid=5476781&rcid=4178379 [00:40:56] !ignore meta.wikimedia user MiszaBot [00:40:57] Rule added. [06:04:07] ooh, our Athens office expands ^^ [08:48:42] database is locked for write access? [08:53:57] matanya: where and what error message did it give? [08:54:13] peachey|laptop__: commons [08:54:30] "The database is currently locked for write access;" [08:54:34] using api [08:55:17] matanya: its working for me atm [08:55:21] (not using the api) [08:55:41] https://commons.wikimedia.org/w/index.php?title=User%3APeachey88&diff=96203489&oldid=55474432 [08:56:10] now it works for me too [08:56:20] maybe a hecap in the db [08:56:53] maybe so. (tested without api, saved edit on commons) [08:57:11] there was some scapping about, perhaps that had an effect for a second or two [14:11:59] are there database problems at the moment? i just got promted http://pastebin.com/qZBPNB0i twice when trying to save an edit (my third attempt was successful). [14:12:24] /s/promted/prompted [17:36:26] greg-g: It's not looking like we're going to get bug 39653 done any time soon, does it? [17:37:43] anomie: what's blocking? Is it just me sending out the notice? [17:37:55] I lost track of that one on Friday [17:38:01] greg-g: Review on 62612 [17:38:44] oh, right [17:38:58] I missed that one now since it wasn't a comment on bug 39653 [17:39:13] guh, how much have you pinged people? [17:39:24] * anomie maybe should mark bug 47235 as blocking 39653 [17:39:41] Not much, don't know who to ping [17:41:14] anomie: the people you added all had previous commits on the codebase, right? I'd pick one randomly and ask nicely [17:41:21] I think I'll ping Krinkle, since https://gerrit.wikimedia.org/r/#/c/62612/ is basically a JavaScript/jQuery change [17:41:27] * greg-g nods [17:41:39] I was going to randomly select Brion, but, that works too :) [17:42:10] * brion hides [17:42:19] :) [17:46:29] hey greg-g: since test isn't for pre-deployment testing anymore, and our labs instance is currently hosed, any advice on testing Echo before we push it to en.wiki today? [17:48:44] heh, mediawiki.org? [17:48:48] kaldari: ^ [17:49:25] is it possible for us to deploy to mediawiki.org and not en.wiki? [17:50:11] if en wiki runs one version of mw and mediawiki.org runs another, then yes [17:50:23] wher eby version I mean 1.21wmfnn [17:50:35] if they run the same version, then... not so much. [17:50:59] well wait, maybe you can... uh. is there a configuration var that controls behavior? [17:51:19] cause in that case yes you could [17:51:50] what apergos said [17:52:42] apergos: huh? [17:52:52] I mean to the last part :) [17:52:57] * greg-g notes it'd be cool for extensions/new features to have $wgEnableWhateverBeta and $wgEnableWhatever [17:52:57] well an wgSomething [17:53:02] yeah [17:53:16] Echo isn't beta it's very much in use on en.wiki [17:53:45] but luckily enwiki's on a different version currently, so we're good [17:53:47] right, I was just referring to the idea that Erik mentioned last week on one of the lists of beta opt-in stuff [17:54:57] being able to push different versions of an extension would be nice, currently don't see how that's possible unless you are lucky enough to have different mw versions. [17:55:55] so where is the common dir on tin? [17:56:45] apergos: oh yeah, that would be very cool [17:56:59] /a/ something [17:57:35] greg-g: the instructions at https://wikitech.wikimedia.org/wiki/How_to_deploy_code#Step_2:_get_the_code_on_tin don't actually work :P [17:58:02] I'll update them [17:58:08] kaldari: :) thanks [17:58:23] I did the dumb s/fenari/tin/g thing on that page ;) [17:58:35] ah I didn't look at that. at all :- [17:58:37] D [17:58:41] only the tin page [17:59:28] is this Echo deployment the first tin deployment? [17:59:57] which would really suck as this is a fairly high risk deployment to start with [18:00:23] kaldari: yeah, I know, I'm sorry about that. :/ [18:00:26] but art least it will be a good test case :) [18:00:49] Reedy: ^^ mind keeping an eye on kaldari, since you've already sync'd out some stuff on tin (I believe) [18:00:55] :) [18:01:00] hmm I guess it is [18:16:14] greg-g, apergos: there is one problem with pushing just to mediawiki.org though, all of our new messages will be broken until the scap. There's isn't any way to update the messages just for wmf4 is there? [18:16:29] hrmm, not that I know of [18:16:37] tbh I have no idea how the l10n stuff works any more [18:16:38] 0 [18:16:42] :) [18:17:00] probably would have to dig aorund in some scripts to find out where the actual rsync is [18:17:07] meh [18:17:34] I tried building a script just for updating i18n caches, but could never get it to actually work [18:19:10] hmm, I guess we have two choices: temporarily break messages on mediawiki.org, or push to en.wiki without testing. I think we should probably go with the first one :) [18:19:22] :) [18:20:27] Is this still up to date: https://wikitech.wikimedia.org/wiki/How_to_do_a_configuration_change#Change_wiki_configuration? [18:24:21] running sync-dir now [18:25:03] bsitu1: no, there are mentions of fenari on there [18:25:10] mw1173: ssh: connect to host mw1173 port 22: Connection timed out [18:25:14] checking that one [18:25:24] kaldari: [18:25:50] thanks [18:26:54] !log powercycling frozen mw1173 [18:27:02] Logged the message, Master [18:27:14] greg-g: I tried the same procedure on tin under /a/common, it doesn't seem to work [18:27:26] hrm [18:27:33] Reedy: ^^ [18:27:33] greg-g: are you saying fenari is not deployment host anymore? weeee [18:27:58] mutante: :) as long as we're able to get through today's deploys, that's right :) [18:27:58] Reedy: do i need to change the way i deploy Apache config? [18:28:06] git fetch origin && git diff HEAD origin doesn't show the configuration change [18:28:08] mutante: I think not. That's still on fenari [18:28:09] :D [18:28:17] well then [18:28:35] thanks both [18:28:54] greg-g: that's great:) [18:29:03] new code is on mediawiki.org and running successfully. JS took a few minutes to kick in, but that's normal. [18:29:04] about moving away from fenari in general [18:29:11] yeah [18:29:14] kaldari: awesome [18:29:44] mw1173 is back online [18:30:06] but may have missed a sync now [18:30:09] Reedy: I think bsitu1 is still stuck, maybe? [18:30:17] I think I've probably pulled it in... [18:30:19] bsitu1: git log [18:30:30] Reedy: let me see [18:31:01] Reedy: yeap, see that change [18:31:19] thought I did something wrong, :) [18:31:37] Reedy: did you sync the file? [18:31:40] My fault, sorry [18:31:42] And yes [18:31:47] Reedy: thx [18:33:25] mutante: I'll try re-syncing for mw1173 [18:33:53] Reedy: I should get shell access to answer these stupid simple questions, but, is wikiversions.dat no under /a/ instead of /home/? [18:33:58] mutante: worked that time, thanks! [18:34:08] kaldari: cool, thx [18:34:13] greg-g: Yup [18:34:19] really, I should't get shell access just to asnswer that question :) [18:34:22] thanks [18:36:28] /home/wikipedia/common (fenari) -> /a/common (tin) [18:36:39] right [18:37:17] i'm gonna graceful apaches for bug 47305 [18:37:18] https://wikitech.wikimedia.org/w/index.php?title=Heterogeneous_deployment&diff=70213&oldid=70207 [18:42:36] apaches or apaches and api servers? :-P [18:43:23] greg-g: So I'm assuming that this section should be deleted from the instructions: https://wikitech.wikimedia.org/wiki/How_to_deploy_code#Trying_tin.27s_code_on_testwiki [18:44:01] Reedy: ^ [18:45:59] I'm not quite sure what state srv193 is in [18:53:21] what i gathered from tim's email sounded like you should be able to syc-common on srv193 (for now) to get stuff on testwiki [18:54:10] as mentioned multiple times previously/elsewhere, the mobile team really depends on testwiki before our deployments, so i *really* hope testwiki is still viable (at least until a workable alternative is available) [18:54:23] particuarly since no one has communicated with us about these changes... [18:58:52] awjr: ah yeah, looks like you're right. I missed that part [18:59:37] I didn't actually test syncing to test.wiki though, so you guys can test that one :) [18:59:54] heh [19:00:03] how has everything else been so far kaldari? [19:00:58] not bad [19:01:19] benny's about to try the scap [19:01:31] groovy [19:02:46] I am about to run scap [19:05:53] awjr: let me know how the sync to srv193 goes, and if it works for your situation. [19:06:48] will do greg-g [22:02:45] !log upgrading fundraising civi (d895730067e3325fd630d09196ee81eeba24b5ee --> head) [22:02:53] Logged the message, Master [23:41:53] !log rolling back civi from deploy head to d895730067e3325fd630d09196ee81eeba24b5ee because thank you isn't working [23:42:02] Logged the message, Master [23:42:57] gn8 folks [23:57:13] !log lies -- went from d895730067e3325fd630d09196ee81eeba24b5ee back to head on civi because it was working -- the tester was just broken [23:57:27] Logged the message, Master