[12:32:06] There's a question here on how to increase concurrency for PHP tools https://en.wikipedia.org/w/index.php?title=User_talk%3ACitation_bot&type=revision&diff=909438452&oldid=909417866 [12:39:45] Nemo_bis: perhaps that's better discussed in phabricator? [12:43:32] arturo: I don't know, do you consider it a bug? [12:43:54] maybe there's some documentation which was missed [12:44:08] I don't know, but that kind of technical discussion is what I would expect to happen in phabricator rather than in-wiki [12:45:11] oh well, I have no control on that so I'll just cross fingers [13:43:58] !log tools disabling puppet on tools-checker-03 while testing nginx timeouts T221301 [13:44:03] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [13:44:03] T221301: Toolschecker webservice checks get out of sync likely from timeouts - https://phabricator.wikimedia.org/T221301 [14:12:03] Nemo_bis: I'm trying to follow the discussion there, but it is not making a lot of sense to me. [14:43:33] bd808: essentially, when one user is waiting for a response to a request to https://tools.wmflabs.org/citations/ all the other users appear to be blocked [14:43:49] I have never managed to figure out whether this is expected behaviour [14:46:57] Nemo_bis: from the default lighttpd config at https://wikitech.wikimedia.org/wiki/Help:Toolforge/Web/Lighttpd, it looks like it runs two worker processes? [14:47:07] so I would expect half of other requests to be blocked, but not all of them [14:47:13] Nemo_bis: ah. ok. Probably partly how the tool is written and partly what Lucas_WMDE is pointing out [14:47:23] not sure how lighttpd interacts with the workers though [14:47:48] but raising max-procs (opposite of https://wikitech.wikimedia.org/wiki/Help:Toolforge/Web/Lighttpd#FCGI_Flask_config ) might be worth a try [15:09:13] Lucas_WMDE: thanks; I wasn't sure whether these could be overridden with kubernetes but I'll try to ask the maintainer to try [15:20:17] Nemo_bis: the .lighttpd.conf parts are the same for both grid and kubernetes [15:58:57] !log tools add hpham as user/projectadmin (T228942) [15:59:01] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [15:59:02] T228942: Onboard Hieu Pham to Wikimedia Foundation as SRE in Cloud Services - https://phabricator.wikimedia.org/T228942 [16:18:11] !log tools add phamhi as user/projectadmin (T228942) and delete hpham [16:18:15] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [16:18:15] T228942: Onboard Hieu Pham to Wikimedia Foundation as SRE in Cloud Services - https://phabricator.wikimedia.org/T228942 [17:09:10] !log deployment-prep deploying ores d08fa628aacb82529dbb4be357b68dd55c15fdee [17:09:14] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Deployment-prep/SAL [17:34:54] anomie: what DB table has the namespaces defines? [17:34:58] *Defined? [17:35:24] Or is that a hardcoded setting in MW? [17:37:56] bd808: ^ [17:41:18] Cyberpower678: it's in PHP files [17:41:31] we used to have a "meta" table for this IIRC [17:43:23] (It's not in meta_p.wiki.) [17:43:32] Nemo_bis: so the only way is via the API? [17:45:05] Cyberpower678: the only reliable way that I know of, yes [17:45:18] if you mean namespace *names* [17:45:56] Nemo_bis: yea. I wanted to pull data from the page table and join a namespace table via the namespace column. Oh well. [17:46:19] I was trying to reduce the number of queries I needed to make to get the data I need. [17:46:19] Cyberpower678: may I ask what for? [17:47:02] Nemo_bis: my IABot tool has a URL manager and domain manager for handling how IABot deals with them. The tool provides a complete list of pages any given URL or domain is found on. [17:47:14] Sometimes a page outside of Mainspace is found. [17:47:34] The tool uses the DB to pull that data as it is faster in this instance. [17:48:05] But it has a a failover to the API if the replication DB goes down/away [17:49:25] Right. I don't remember how Magnus solved this. He used to have such a problem in glamtools for image usages [17:50:33] Well it's an easy fix for me. I just wanted to know if there is a table for the DB that I can simply amend the DB query for. [18:34:14] Hi, [18:34:47] anyone know how should I request access to a cloud instance if I need to test a operations/puppet patch? [18:42:22] depending on the patch in question there may already be a relevant place to test it