[00:18:27] DatGuy: that sounds similar to https://phabricator.wikimedia.org/T376849 [09:48:32] !log urbanecm@tools-bastion-13 tools.wikinity Deploy 9743f47 (T331193) [09:48:36] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.wikinity/SAL [11:36:56] Why do I get a 200 response with an empty body from stream Wikimedia? [20:21:53] Please remove this page: [20:22:37] http://medbox.iiab.me/modules/en-cdc/www.cdc.gov/cancer/dcpc/about/leadership/m_white.htm [20:23:09] Please delete this page because it provides an identifible photograph of me and outdated information about me. The page incorrectly indicates that I hold a leadership position at the CDC. I no longer work at this organization (CDC). This information poses a targeted security risk to me personally, because persons who hold leadership positions at [20:23:10] CDC are targets of harassment and political violence in the US. [20:24:36] !help [20:24:36] If you don't get a response in 15-30 minutes, please email the cloud@ mailing list -- https://wikitech.wikimedia.org/wiki/Help:Cloud_Services_communication [20:26:59] Guest84: this is not the right channel for general wiki requests. Also, it seems likely that you can make the edits yourself at en.wikimedia.org and the change will be propagated to iiab when they update. [20:28:06] That page specifically looks like it's from a mirror of the CDC website, not from a Wikimedia project anyway [20:28:40] stw: good point [20:29:42] * andrewbogott did not click [20:41:38] What is the right channel for wiki requests? [20:43:10] Guest84: that website is a mirror of the CDC website [20:43:15] Nothing to do with wikimedia [20:43:45] What is the web address for en.wikimedia.org? [20:44:39] RhinosF1: the specific site they are asking about is hosted by WMCS in a project maintained by Doc James and Timmoody. https://openstack-browser.toolforge.org/project/iiab [20:45:34] Guest84: this guys is your best bet for any changes -- https://meta.wikimedia.org/wiki/User:Doc_James [20:45:38] *guy [20:46:11] bd808: how did you link the project to that domain [20:47:22] RhinosF1: `host medbox.iiab.me` medbox.iiab.me has address 185.15.56.46; `host 185.15.56.46` 46.0-25.56.15.185.in-addr.arpa domain name pointer instance-medbox4-iiab.iiab.wmcloud.org. [20:47:25] Guest84: sorry, the instructions I gave you about wikimedia.org contained a typo and were also wrong :( [20:48:10] bd808: ah right, so nothing I was obviously missing. I was searching the site for any indication of a contact or help option. [20:48:12] I actually already knew that the medbox.iiab.me vanity name was WMCS hosted from past debates about such things [21:50:25] Ladies, gentlemen and others; it appears that I am back. [21:52:01] The 10 year break has done me good, and I have fresh ideas to implement ;) [21:54:04] My first question: Is the Sun Grid Engine still a thing? [21:55:21] a930913: nope. https://wikitech.wikimedia.org/wiki/News/2024_Toolforge_Grid_Engine_deprecation [21:56:13] Ngl, I'm kind of shocked that 2024 is in the title. [21:56:58] it was a very long journey to kill it off [22:01:26] So I have a complex SQL query that I want to turn into a tool. I have it running locally, but it uses the repdb and temporary tables. I presume I can't use both at the same time on the labs, can I? [22:02:11] "labs" was renamed in 2017 because nobody knows which "labs" you mean. [22:02:55] Oh, lol. Uh, toolforge? [22:05:14] The Wiki Replica servers probably do not support a temporary table solution, but that may depend on how temporary. I think you can select into an in-memory table that only exists for the duration of the query, but I may be wrong about that. [22:06:46] The 2017 Wiki Replicas redesign had to kill off the former feature of user maintained tables co-existing with the replicated MediaWiki tables. More at https://phabricator.wikimedia.org/phame/post/view/70/new_wiki_replica_servers_ready_for_use/ [22:06:56] When I tried CREATE TEMPORARY TABLE from the interactive client on the tool's user, it threw a permission error. [22:08:46] I guess I was wrong that you could do an in-memory create temp table then? [22:11:11] That could be the case. I presume it's futile to try replicate the replication in the tools db? I should try and rewrite the query all in memory. [22:12:15] `show grants;` tells me that the wiki replica users can select and show view. [22:12:27] so yeah, no create temp table. [22:15:44] a930913: "I presume it's futile to try replicate the replication in the tools db?" -- probably? You could certainly build a tool that finds some data in the wiki replicas and stores something in toolsdb as a result, but if you are looking to keep a mirror of any non-trivial amount of data toolsdb is probably a bad fit. [22:40:29] Hmm. It takes too much memory, as it needs to record thousands of categories it has checked. I suppose I'll have to use the memory of the application layer, though I may end up hammering the database :/ [22:51:34] a930913: does your workflow need to have immediate answers? If not you can think about stretching the runtime out to be nicer to shared services. [22:52:33] but generally you should be able to go hard against the *.analytics replicas and trust that the server is tuned to keep you from starving everyone out. [22:53:39] staying single threaded for queries is appreciated however [22:54:32] The idea would be to provide an immediate answer there and then. [22:55:31] The TL;DR> is a tool that traverses the category tree to give you the path from a category to its subcategory. [22:57:06] So a breadth first search which can go quite deep, can take a fair chunk of compute. [23:02:28] https://www.mediawiki.org/wiki/Wikidata_Query_Service/Categories might also be useful for working with the category graph btw [23:02:37] it’s used by the deepcat: search keyword https://www.mediawiki.org/wiki/Help:CirrusSearch#Deepcategory