[02:15:20] !log LocalisationUpdate completed (1.19) at Sat Apr 21 02:15:20 UTC 2012 [02:15:44] Logged the message, Master [02:29:40] !log LocalisationUpdate completed (1.20wmf1) at Sat Apr 21 02:29:40 UTC 2012 [02:29:42] Logged the message, Master [05:06:16] http://commons.wikimedia.org/wiki/File:St_Francis_de_Sales_Regional_College_(3).jpg isn't showing http://upload.wikimedia.org/wikipedia/commons/thumb/1/12/St_Francis_de_Sales_Regional_College_%283%29.jpg/800px-St_Francis_de_Sales_Regional_College_%283%29.jpg gets an error [05:06:31] Error creating thumbnail: convert: Output file write error --- out of disk space? `/tmp/transform_1abd4ae-1.jpg' @ error/jpeg.c/EmitMessage/235. [05:09:00] Bidgee|1: still? [05:09:16] i can't get it to fail [05:10:15] Bidgee|1: ping [05:14:33] Seems to have corrected itself now [05:23:08] probably the tmp dir was full, we clear out anything older than 15 mins but still [05:25:06] apergos: what's full mean? i only saw some generic (not per FS) disk metrics in ganglia. but maybe i didn't look hard enough [05:25:31] full means the root filesystem [05:25:39] full means 0 bytes left [05:25:50] none of them had full disks recently per ganglia [05:25:52] but given that we clear stuff out in cron every 5 minutes [05:26:05] not even close [05:26:10] on the scalers? [05:26:13] we see it all the time [05:26:48] http://ganglia.wikimedia.org/latest/?r=hour&cs=&ce=&m=disk_free&s=by+name&c=Image+scalers+pmtpa&h=&host_regex=&max_graphs=0&tab=m&vn=&sh=1&z=small&hc=4 [05:26:58] assuming we're talking about pmtpa [05:27:33] i don't see an easy way (short of looking at the preseed partman recipes) to see what kind of disk layout they have [05:27:44] to know if /tmp is it's own filesystem [05:27:53] well we have been dealing with this issue for quite some time now [05:27:58] and it's not its own filesystem [05:28:01] and I guarantee you [05:28:29] that we have seen these boxes fil up those dirs over and over, which is why we clear them out as often as we do [05:29:01] right, it certainly rings a bell as a chronic (but not quite recent) problem. and the automated cleaning rings a bell too [05:29:02] I can't actually get to ganglia right now, my bandwidth is bad enough that if I download anything then udp fails (= dns fails for me) [05:30:17] well ganglia says they've all had ~60+ GB free for the entire last hr [05:31:00] that's likely total and not root only [05:31:07] root only isn't 60, it's much smaller [05:31:16] like 6 or 7 [05:31:33] oh ;( (/me points back to "don't know the disk layout") [05:31:49] uh huh [05:31:56] theres a rough layout of it on wikitech i believe [05:32:01] this is why I was sure(cause I"ve had to stomp on those hosts many a time) [05:32:02] but its slightly outdated [05:32:13] hrmmmm, i wonder what "maximum disk space used" does [05:32:50] I know chad or reedy ploted some notes down someone, maybe wikitech-l, for what they would like the layout to be in the future, possibly wikitech [05:33:39] I know what I would like, and that is to have reasonable size root partitions in the future, and to have /var and /tmp on their own partitions with reasonable space for those too. someday [05:33:49] that's an across the bboard thing [05:34:00] this one tells a totally different story: http://ganglia.wikimedia.org/latest/?r=hour&cs=&ce=&m=part_max_used&s=by+name&c=Image+scalers+pmtpa&h=&host_regex=&max_graphs=0&tab=m&vn=&sh=1&z=medium&hc=2 [05:34:09] all of them reached 100% in the last hr [05:34:23] but while I grumble about it, I also have my todo list which doesn't have that on it, and most items on it are very overdue [05:34:26] * jeremyb would love some per filesys metrics! [05:35:13] normally we get told by nagios, but I didn't see any of that just now looking in the whine channel [05:35:33] yeah i don't see it either [05:35:43] so it's 8:35 am here on a Saturday and while I was happy to chime in and explain the likely cause, [05:35:58] I'm going to not do much work-related today if I can get away with it [05:36:23] are you going to do something fun? ;) [05:36:28] well hmm [05:36:46] I have various errands and stuff to do (clean house. buy cat food. get groceries. etc) [05:36:54] * jeremyb is going to an editathon at that time tomorrow [05:36:56] so that'snot exactly fun, but it does need doing [05:37:20] also we had a police raid in our neighborhood yesterdya [05:37:21] well enjoy ;) [05:37:23] lasted half the day [05:37:30] oh, we had one here too [05:37:49] today is a concert which was supposed to happen with the support of one of the raided places [05:37:51] https://mayfirst.org/fbi-seizes-mfpl-server-attack-anonymous-speech [05:38:07] I guess it's going to go on anyways, I had planned to go so I guess we'll see how that turns out [05:38:17] enjoy [05:38:28] of f*ck they hit riseup? really? [05:39:09] they seized it for mixmaster? wtf [05:39:27] jesus [05:39:42] afaik it is not illegal to operate a mixmaster server [05:39:45] not by a long shot [05:41:37] no english text, sorry, but here you can see some photos of the cops at work [05:41:39] http://athens.indymedia.org/front.php3?lang=el&article_id=1393554 [05:41:46] they blocked off many square blocks [05:42:10] it was virtually impossible to circulate anywhere in the neighborhood [05:42:46] they were after two occupied buildings (no reason given, the whole thing started at 6 am and went til at least 2 pm) [14:33:43] how long does it typically take for a post to propagate on wikitech-l when sent out via gmane? i only post there very very occasionally, and i forget if propagation is relatively quick or not. [14:34:38] i posted about 20 minutes ago and don't see my message in http://news.gmane.org/gmane.science.linguistics.wikipedia.technical/ [14:46:23] Emw, maybe your address is not subscribed to the list? [14:48:06] that'd do it [14:48:23] thanks, will try that in a little while [17:36:10] Joan: did you get gerrit data extraction figured out? [18:03:35] umm [18:04:45] oh nm [18:54:59] jeremyb: I don't think so. [18:55:12] I saw something from Platonides about a manganese DB. [18:55:12] But I don't think that's publicly available. [18:55:16] I imagine it needs to have views added to it. [18:55:26] If it's going to be public at all. [18:55:38] I guess the broader question is "how are you supposed to access gerrit data?" [18:55:45] Does it have an API or something? [19:34:26] Joan: you can just fetch most data with git itself i think. (and i tested) [19:34:30] Joan: poke me when you're back [19:39:10] Joan, manganese is the name of the server [19:45:52] jeremyb: Are there docs on that? [19:46:59] jeremyb: My only issue was that Erik's message seemed to be looking for volunteers, but didn't seem to explain how to get the data. [19:50:26] oh, sorry walked away again [19:50:37] I've been in and out. [19:50:44] no i mean i did [19:50:53] Right. [19:50:54] Same. [19:51:04] The question was about data access docs. [19:51:08] Or data access methods. [19:51:11] Joan: idk. there's gerrit docs which may cover it or not. or you can just use the suggestion from the email. which i now tested [19:51:28] Is that how you're supposed to get data? [19:51:29] 1 sec [19:51:29] Over SSH? [19:52:19] Regardless of other access methods, high-level (SQL) access to the data is better... [19:52:32] Though I don't know about SSH. [19:53:57] > [19:53:58] Here's the technique, which works if you've got a valid gerrit account: [19:53:58] ssh -p 29418 gerrit.wikimedia.org gerrit query 'status:open [19:53:58] project:^mediawiki/.*' | grep rowCount [19:53:59] > [19:54:26] sorry, in person distraction [19:54:37] You don't have to apologzie. [19:54:40] apologize [19:54:40] I want {{#formatdate:{{{2}}}|dmy}} to generate language specific output how can I do this? [19:54:58] ToAruShiroiNeko: Do the docs indicate that it supports language specific output? [19:55:01] It might be an open bug. [19:55:06] I'd check the docs, then Bugzilla. [19:55:29] It must support language. [19:55:34] Isn't it based on user preference or something? [19:55:38] And then has some default for anons? [19:55:44] I think this is all documented. [19:56:57] is there any info you'd want to have (or which could be had from your hypothetical table view) besides 1) the logs for the various projects, tags, branches, and changes (with all individual patch sets) 2) the reviews/verifications/comments on those changes [19:57:02] ? [19:57:17] Joan the problem is auto-date translate is broken [19:57:27] or is insufficient to what I need to do [19:57:52] can I force that code or some other code to reurn date in dmy format and in spanish [19:59:11] You're trying to force a language? [19:59:13] Did you read the docs? [19:59:25] yes {{#formatdate:{{{2}}}|dmy}} doesnt have an option [19:59:30] for that [19:59:34] it depends on user selection [20:00:05] {{#formatdate:{{{2}}}|dmy}} doesnt seem to work at all for spanish I am told [20:02:03] Right. [20:02:06] So you can look for and/or file a bug. [20:02:18] I'd rather not [20:02:25] I want a quick solution [20:02:42] not some bug request that will take years to resolve [20:14:57] surprise surprise I've heard a bug report about that [20:15:08] can't dig now [20:16:17] for the purpose of this template forced translation would work better [20:16:39] all the content is un spanish anyways [22:14:12] apergos: that anon email relayer was seize because it was linked to the bomb threats at pittsburg uni. (if you enever ended up finding out)