[01:39:52] RECOVERY - wiki.mlpwiki.net - reverse DNS on sslhost is OK: rDNS OK - wiki.mlpwiki.net reverse DNS resolves to cp10.miraheze.org [02:42:07] [02mw-config] 07Universal-Omega opened pull request 03#3846: add more file formats to wgFileExtensions - 13https://git.io/JO6hI [02:43:01] miraheze/mw-config - Universal-Omega the build passed. [04:43:51] PROBLEM - wiki.mlpwiki.net - reverse DNS on sslhost is CRITICAL: rDNS CRITICAL - wiki.mlpwiki.net reverse DNS resolves to 192-185-16-85.unifiedlayer.com [04:44:07] PROBLEM - cp11 Current Load on cp11 is CRITICAL: CRITICAL - load average: 8.66, 5.08, 2.28 [04:46:06] PROBLEM - cp11 Current Load on cp11 is WARNING: WARNING - load average: 1.27, 3.46, 2.02 [04:48:05] RECOVERY - cp11 Current Load on cp11 is OK: OK - load average: 0.38, 2.42, 1.80 [05:25:10] [02miraheze/services] 07MirahezeSSLBot pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOiG6 [05:25:12] [02miraheze/services] 07MirahezeSSLBot 03a3f89c3 - BOT: Updating services config for wikis [10:20:28] PROBLEM - wiki.abyogsms.org.in - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for wiki.abyogsms.org.in could not be found [10:45:51] RECOVERY - wiki.mlpwiki.net - reverse DNS on sslhost is OK: rDNS OK - wiki.mlpwiki.net reverse DNS resolves to cp11.miraheze.org [12:23:57] [02miraheze/dns] 07Reception123 pushed 031 commit to 03master [+1/-0/±0] 13https://git.io/JOPsE [12:23:58] [02miraheze/dns] 07Reception123 037becd21 - add ferrandalmeida.family zone [12:27:05] [02miraheze/ssl] 07Reception123 pushed 031 commit to 03master [+1/-0/±1] 13https://git.io/JOPsj [12:27:07] [02miraheze/ssl] 07Reception123 0354d4187 - add www.ferrandalmeida.family cert [12:30:53] [02mw-config] 07Reception123 commented on pull request 03#3846: add more file formats to wgFileExtensions - 13https://git.io/JOPGp [12:34:08] [02miraheze/ssl] 07Reception123 pushed 031 commit to 03master [+1/-0/±1] 13https://git.io/JOPZA [12:34:09] [02miraheze/ssl] 07Reception123 03ef2f516 - add wiki.creaturathegame.com cert [12:41:14] paladox: hi. Any idea what we'd do about https://phabricator.miraheze.org/T7174 ? [12:41:15] [ ⚓ T7174 Wikibase Concept URI Links Do Not Work ] - phabricator.miraheze.org [12:41:42] why is it doing https://batman.miraheze.org/entity/Q1? [12:41:43] [ 404 Not Found ] - batman.miraheze.org [12:41:48] but no :( [12:43:07] my guess is that it's not working because the URL is directly /entity [12:43:08] but not sure what we need to do to fix that [12:43:24] Reception123, it seems like the wiki URL structure is wrong, and WikiBase is missing the /wiki/ prefix for the entities [12:43:33] lol same thing! [12:44:30] guess it's https://github.com/miraheze/mw-config/blob/master/Wikibase.php#L25 [12:44:31] PROBLEM - cp3 Puppet on cp3 is CRITICAL: CRITICAL: Failed to apply catalog, zero resources tracked by Puppet. It might be a dependency cycle. [12:44:31] [ mw-config/Wikibase.php at master · miraheze/mw-config · GitHub ] - github.com [12:44:53] paladox: that's from the sidebar in tools [12:45:12] screenshot as i cannot see anything related in tools [12:46:10] if you do https://www.wikidata.org/entity/Q1 it redirects [12:46:15] so not sure why it doesn't for us? [12:49:14] looks like it needs a redirect [12:50:45] https://github.com/wikimedia/puppet/blob/e1e13a59de3021afaa43c31745abbe348a93017d/modules/mediawiki/files/apache/sites/wikidata-uris.incl#L41 [12:50:46] [ puppet/wikidata-uris.incl at e1e13a59de3021afaa43c31745abbe348a93017d · wikimedia/puppet · GitHub ] - github.com [12:52:32] PROBLEM - mon2 Puppet on mon2 is CRITICAL: CRITICAL: Failed to apply catalog, zero resources tracked by Puppet. It might be a dependency cycle. [12:52:38] [02miraheze/puppet] 07paladox pushed 031 commit to 03paladox-patch-2 [+0/-0/±1] 13https://git.io/JOPCp [12:52:40] [02miraheze/puppet] 07paladox 031462d43 - Redirect Redirect /entity/* to /wiki/Special:EntityData/* [12:52:41] [02puppet] 07paladox created branch 03paladox-patch-2 - 13https://git.io/vbiAS [12:52:43] [02puppet] 07paladox opened pull request 03#1743: Redirect Redirect /entity/* to /wiki/Special:EntityData/* - 13https://git.io/JOPCh [12:53:43] PROBLEM - cp12 Puppet on cp12 is CRITICAL: CRITICAL: Failed to apply catalog, zero resources tracked by Puppet. It might be a dependency cycle. [12:54:03] [02puppet] 07paladox closed pull request 03#1743: Redirect Redirect /entity/* to /wiki/Special:EntityData/* - 13https://git.io/JOPCh [12:54:05] [02miraheze/puppet] 07paladox pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOPWT [12:54:06] [02miraheze/puppet] 07paladox 03d078b9b - Redirect Redirect /entity/* to /wiki/Special:EntityData/* (#1743) [12:54:08] [02puppet] 07paladox deleted branch 03paladox-patch-2 - 13https://git.io/vbiAS [12:54:09] [02miraheze/puppet] 07paladox deleted branch 03paladox-patch-2 [12:54:34] paladox: thanks! so then it was just supposed to be a 301 redirect [12:55:15] PROBLEM - cp10 Puppet on cp10 is CRITICAL: CRITICAL: Failed to apply catalog, zero resources tracked by Puppet. It might be a dependency cycle. [12:56:45] ?[1;31mError: Could not retrieve catalog from remote server: Error 500 on SERVER: Server Error: Evaluation Error: Error while evaluating a Function Call, (/etc/puppetlabs/puppet/ssl-cert/certs.yaml): could not find expected ':' while scanning a simple key at line 1153 column 6 (file: /etc/puppetlabs/puppet/environments/production/modules/mediawiki/manifests/nginx.pp, line: 4, column: 17) on node mw10.miraheze.org?[0m [12:57:14] [02miraheze/ssl] 07paladox pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOPWQ [12:57:15] [02miraheze/ssl] 07paladox 03a952be5 - Fix [12:57:35] whoops, thanks for fixing that too :) [12:58:23] [02mw-config] 07dmehus opened pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [12:59:17] Redirects now [12:59:27] miraheze/mw-config - dmehus the build passed. [12:59:31] paladox, how do you add additional commits in your fork to an existing PR? [12:59:33] great! [12:59:47] I thought it would add them all automatically when I hit "compare and pull requests" [12:59:49] you just click edit on a file then you can view the file tree [12:59:50] dmehus: you just edit the files on the branch [13:00:26] paladox and Reception123, ah, thanks. Yeah, I've edited all the files in my branch, but not sure how to get the commits added to the PR [13:01:08] https://github.com/miraheze/mw-config/compare/master...dmehus:master [13:01:09] [ Comparing miraheze:master...dmehus:master · miraheze/mw-config · GitHub ] - github.com [13:02:08] https://github.com/dmehus/mw-config/blob/patch-10/extension-list ergh my change isn't there. Did it add it as a different branch? [13:02:08] [ mw-config/extension-list at patch-10 · dmehus/mw-config · GitHub ] - github.com [13:02:31] Reception123, oh thanks [13:02:50] dmehus: ah yes, I see. You accidentally did it on the master branch [13:02:55] instead of editing the files on the patch-10 branch [13:03:01] wait that's going to create a new PR [13:03:09] oh [13:03:12] yeah, that's not good then since you got the wrong branch [13:03:14] RECOVERY - cp10 Puppet on cp10 is OK: OK: Puppet is currently enabled, last run 49 seconds ago with 0 failures [13:03:23] before you had to select the patch-10 branch and only then edit the files [13:03:28] ergh there's got to be a way to fix this [13:03:42] dmehus: yeah, just copy what you did on master to the patch-10 branch [13:03:44] RECOVERY - cp12 Puppet on cp12 is OK: OK: Puppet is currently enabled, last run 5 seconds ago with 0 failures [13:03:46] and then the PR will update [13:03:53] so in https://github.com/dmehus/mw-config/tree/patch-10 [13:03:54] [ GitHub - dmehus/mw-config at patch-10 ] - github.com [13:04:00] can't I add those ones on the master branch to my PR? [13:04:18] well it would be tricky, it's easier to just copy/paste the lines [13:04:29] can I do it without copy/pasting, by merging my commits from master to patch-10? [13:04:51] you can try yeah [13:04:55] in theory it should work [13:05:22] okay, I can try that, how would I do that though? [13:06:04] there should be a notice saying "This branch is ahead of" and should propose a PR to change that [13:06:26] ah [13:06:29] yeah [13:06:35] RECOVERY - cp3 Puppet on cp3 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [13:06:55] [02miraheze/landing] 07translatewiki pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOP8V [13:06:56] [02miraheze/landing] 07translatewiki 037b31030 - Localisation updates from https://translatewiki.net. [13:06:57] [ Main page - translatewiki.net ] - translatewiki.net [13:06:58] [02miraheze/ManageWiki] 07translatewiki pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOP8w [13:06:59] [02miraheze/ManageWiki] 07translatewiki 03666e4ae - Localisation updates from https://translatewiki.net. [13:07:00] [ Main page - translatewiki.net ] - translatewiki.net [13:07:01] [02miraheze/MirahezeMagic] 07translatewiki pushed 031 commit to 03master [+0/-0/±4] 13https://git.io/JOP8r [13:07:02] [02miraheze/MirahezeMagic] 07translatewiki 03b8aae98 - Localisation updates from https://translatewiki.net. [13:07:03] [ Main page - translatewiki.net ] - translatewiki.net [13:08:03] miraheze/MirahezeMagic - translatewiki the build passed. [13:08:09] miraheze/landing - translatewiki the build passed. [13:08:18] miraheze/ManageWiki - translatewiki the build passed. [13:08:33] Reception123, the confusing part is what do I click first to change miraheze/mw-config-->master --> dmehus/mw-config-patch10 etc. etc. without it dropping the latter half of the equation [13:08:38] if that makes sense [13:08:57] [02miraheze/mediawiki] 07Reception123 pushed 032 commits to 03REL1_35 [+1/-0/±2] 13https://git.io/JOP4J [13:08:59] [02miraheze/mediawiki] 07Reception123 03f0af02b - Install PageAssessments ext. T7180 [13:09:00] [02miraheze/mediawiki] 07Reception123 034f9f50a - +ignore dirty [13:09:10] dmehus: hmm yeah, it is a bit confusing which is why I initially suggested copy/pasting instead [13:12:55] Reception123, I'd rather learn how to fix it rather than redo it. Maybe a screenshot would help and you can tell me what I need to click first [13:13:16] let me see, though I'm no git expert heh [13:13:19] https://usercontent.irccloud-cdn.com/file/kvJ3iLFa/2021-04-22%2006.12.12%20github.com%2072bca528c744.jpg [13:13:53] PROBLEM - test3 MediaWiki Rendering on test3 is CRITICAL: HTTP CRITICAL: HTTP/1.1 500 Internal Server Error - 324 bytes in 0.029 second response time [13:14:13] dmehus: that looks right but then when I actually go to that it seems to only propose to merge one of the files [13:14:24] oh I think I figured it out [13:14:26] on my PR [13:14:35] I can change the destination to dmehus/mw-config/master [13:14:51] yeah, but I think that would conflict [13:15:07] because your master would be behind the actual mw-config master, since many things have happened on the original mw-config since you forked your dmehus/mw-config repo [13:15:13] oh no nevermind, yeah has to be in your repository [13:15:28] okay [13:15:48] yeah, git isn't easy at all. After about 5 years I still barely get many things [13:15:49] so it's probably better just to cancel this PR and redo that one comit [13:15:52] commit* [13:16:01] well you don't even need to "cancel" the PR [13:16:02] and merge from my master branch to mw-config? [13:16:10] all you'd do is go to the right branch and then copy your master changes to patch-10 [13:16:34] that's more work though since only one commit is in patch-10 and three are in master [13:16:47] so I'd rather initiate a PR from my master branch if possible [13:17:05] yeah but the issue with initiating from master is that master won't be up to date with miraheze/mw-config [13:17:09] unless you forked your repo very recently [13:17:30] yeah, it says "This branch is 3 commits ahead, 374 commits behind miraheze:master." [13:18:08] well I did hit "edit and fork this file" or whatever [13:18:11] so that's odd [13:18:21] hm [13:18:23] I meant my master [13:18:27] not miraheze/master [13:18:37] PROBLEM - test3 Puppet on test3 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:18:47] or you're saying my master is behind my patch-10? [13:19:01] yes [13:19:11] because patch-10 was forked directly from mw-config/master in its current state [13:19:17] your master was only forked whenever you forked mw-config [13:19:24] ah [13:19:36] another reason I should be deleting my fork after patch [13:19:37] heh [13:20:20] I guess I'll redo it then [13:20:41] RECOVERY - mon2 Puppet on mon2 is OK: OK: Puppet is currently enabled, last run 52 seconds ago with 0 failures [13:21:09] probably easier yeah [13:21:49] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:22:36] * dmehus makes a note to himself to note what branch he is in when editing in one of his forks [13:22:53] miraheze/mw-config - dmehus the build passed. [13:23:24] * Reception123 is annoyed that test3 doesn't want to reclone /srv/mediawiki/config [13:23:37] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:24:38] miraheze/mw-config - dmehus the build passed. [13:24:56] and ironically it's also because I had to delete the /srv/mediawiki/config to get rid of the changes ahead of mw-config [13:25:16] (since I'm preparing for switching the forks and want to make sure test3 is on the same "page" as the other mw*s) [13:27:21] urgh [13:27:22] Notice: /Stage[main]/Mediawiki/Git::Clone[MediaWiki core]/Exec[git_pull_MediaWiki core]/returns: error: insufficient permission for adding an object to repository database /srv/mediawiki/w/.git/modules/skins/Metrolook/objects [13:28:02] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:28:13] Reception123, oh [13:28:26] PR should be done now [13:28:37] dmehus: thanks! won't be able to test it until I figure out test3 though [13:29:08] miraheze/mw-config - dmehus the build has errored. [13:29:18] oh [13:29:27] there seems to be an error with my last commit [13:29:27] and your PR seems to not be valid PHP either heh [13:29:30] will try and resolve [13:29:43] sounds good, meanwhile I'll try to resolve test3 [13:29:56] RECOVERY - test3 MediaWiki Rendering on test3 is OK: HTTP OK: HTTP/1.1 200 OK - 20764 bytes in 0.326 second response time [13:30:00] oh heh, I did it. Just used the pala dox method [13:30:34] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:30:39] RECOVERY - test3 Puppet on test3 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [13:31:41] not sure what the paladox method is with your issue, but wth I just copied the existing code from AJAXPoll, and it's still not working. Let me look again [13:31:43] miraheze/mw-config - dmehus the build has errored. [13:31:47] dmehus: it's not the comma, you're actually allowed to not have a comma for the last line in PHP. the issue is you're missing a closing array [13:31:47] https://github.com/miraheze/mw-config/pull/3847/files#diff-2255ee817629e0892a49d90cb22f97f34dacec6512b327ee446806800f91ece0R2447 [13:31:48] [ Add PageAssessments extension #T7180 by dmehus · Pull Request #3847 · miraheze/mw-config · GitHub ] - github.com [13:32:11] [02mw-config] 07Reception123 reviewed pull request 03#3847 commit - 13https://git.io/JOPEy [13:33:23] oh nevermind [13:33:30] I figured it out yeah [13:33:34] the comma wasn't it [13:33:42] I need an extra set of square brackets [13:33:45] yup [13:37:35] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:38:31] [02mw-config] 07dmehus reviewed pull request 03#3847 commit - 13https://git.io/JOPzO [13:38:37] miraheze/mw-config - dmehus the build passed. [13:40:25] PROBLEM - mw11 Puppet on mw11 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 2 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:41:01] The extension PageAssessments failed to be installed. [13:41:04] hmm, not sure why it doesn't like it [13:41:05] PROBLEM - mw8 Puppet on mw8 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:41:07] PROBLEM - jobrunner3 Puppet on jobrunner3 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 2 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:41:43] PROBLEM - jobrunner4 Puppet on jobrunner4 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:42:10] PROBLEM - mw9 Puppet on mw9 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:42:12] PROBLEM - mw10 Puppet on mw10 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [13:42:44] dmehus: oh I know, not sure where you got those SQL paths [13:42:54] usually they're in /sql yeah, but for this extension the directory is called db https://github.com/wikimedia/mediawiki-extensions-PageAssessments/tree/master/db [13:42:54] [ mediawiki-extensions-PageAssessments/db at master · wikimedia/mediawiki-extensions-PageAssessments · GitHub ] - github.com [13:43:26] Reception123, I just copied the paths from AJAXPoll and inserted the db table names from Extension:PageAssessments [13:43:39] I DMed you if that was right, but guess you didn't see [13:43:57] ah, but also why those DB titles? [13:44:18] dmehus: I didn't get anything, must've been during my timeout [13:44:25] https://www.mediawiki.org/wiki/Extension:PageAssessments says those are the table names, no? [13:44:26] [ Extension:PageAssessments - MediaWiki ] - www.mediawiki.org [13:44:45] Reception123, ah yeah [13:44:58] for the db tables to add, do I just follow: [13:44:58] ``` [13:44:58] 'install' => [ [13:44:58] 'sql' => [ [13:44:58] 'ajaxpoll_info' => "$IP/extensions/AJAXPoll/sql/create-table--ajaxpoll_info.sql", [13:44:58] 'ajaxpoll_vote' => "$IP/extensions/AJAXPoll/sql/create-table--ajaxpoll_vote.sql" [13:44:58] ], [13:44:59] ``` [13:44:59] and replace those table names with the extension's table names? [13:45:06] yeah [13:45:09] see https://github.com/wikimedia/mediawiki-extensions-PageAssessments/tree/master/db [13:45:09] [ mediawiki-extensions-PageAssessments/db at master · wikimedia/mediawiki-extensions-PageAssessments · GitHub ] - github.com [13:45:17] (don't add the patch though, only the main tables) [13:45:43] oh [13:45:51] so for the first it would be 'page_assessments_projects' => path/addProjectsTable.sql [13:45:57] (with the correct syntax) [13:46:01] oh [13:46:25] is the path part still correct /$IP/extensions/PageAssessments/sql/? [13:47:27] dmehus: no, because this extension (weirdly) uses /db instead of /sql [13:47:29] I see there's also a maintenance script, "The purgeUnusedProjects.php maintenance script can be run to remove projects from the page_assessments_projects table that are no longer used by any assessments." Do we need to download that maintenance script separately, or will that be in the package when you clone MediaWiki repo? [13:47:35] oh [13:47:51] so basically the same except change /sql to /db? [13:48:10] dmehus: that's included when I added the extension submodule [13:48:26] Reception123, ah okay, cool :) [13:48:29] the SQL is too, all you're doing is getting it to automatically create on the wiki where it's enabled [13:48:35] yeah [13:48:37] dmehus: well yes but why did you use those weird file names? [13:48:48] "create-table--page_assessments.sql" should just be addProjectsTable.sql [13:48:52] but oh [13:48:53] as it's just the name of the SQL file [13:49:07] oh create-table is part of AJAXPoll's file names? [13:49:24] that's weird, I though it was part of the SQL command to add the table lol [13:49:56] * dmehus thinks there should be standardization in extension database table file names :P [13:50:28] yeah, it's not the same here [13:50:30] ok fixing now then [13:50:33] thanks [13:53:46] [02mw-config] 07dmehus synchronize pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [13:53:58] RECOVERY - mw10 Puppet on mw10 is OK: OK: Puppet is currently enabled, last run 53 seconds ago with 0 failures [13:54:07] RECOVERY - mw9 Puppet on mw9 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [13:54:25] RECOVERY - mw11 Puppet on mw11 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [13:54:39] RECOVERY - mw8 Puppet on mw8 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [13:54:52] miraheze/mw-config - dmehus the build passed. [13:55:01] that will do :) [13:55:56] Reception123, heh cool, thanks for your help :) [13:56:35] no problem. Now to figure out how it works to test it out [13:58:31] Reception123, yeah I think you need to define the projects with: [13:58:31] ```{{#assessment:Medicine|A|Low}} [13:58:31] {{#assessment:Biology|B|High}}``` [13:58:31] then add that to a page, or template, then a special page should show all the pages for that project and their assessment value [13:58:51] yeah, I tried that but it didn't seem to show on the special page [13:58:51] like PageTriage, it's very enwiki specific, but there's potentially some Miraheze use cases [13:58:57] oh [13:59:09] was it a talk page? [13:59:18] as it might be disabled on non-talk pages by default [13:59:31] ah, ok [13:59:38] the thing is talkpages are currently broken on test3 heh [13:59:40] let me try in another NS [13:59:44] oh lol [14:00:03] We could set `$wgPageAssessmentsOnTalkPages` to false [14:00:13] do we need to add that config to ManageWiki? [14:00:13] [02miraheze/services] 07MirahezeSSLBot pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOPaz [14:00:14] * Reception123 will likely drop test3 and get a clean start after 1.36 [14:00:14] [02miraheze/services] 07MirahezeSSLBot 03b02a385 - BOT: Updating services config for wikis [14:00:30] dmehus: got it to work [14:00:33] Reception123, 👍 to dropping and recreating test3wiki yeah [14:00:33] just used the Project NS instead [14:00:37] Reception123, ah ok [14:00:37] Reception123: just make sure to only drop test3 xD jk [14:00:48] [02mw-config] 07Reception123 closed pull request 03#3847: Add PageAssessments extension #T7180 - 13https://git.io/JOPlv [14:00:50] [02miraheze/mw-config] 07Reception123 pushed 031 commit to 03master [+0/-0/±4] 13https://git.io/JOPaK [14:00:51] [02miraheze/mw-config] 07dmehus 03c387b69 - Add PageAssessments extension #T7180 (#3847) [14:00:53] [02mw-config] 07Reception123 synchronize pull request 03#3845: Merge 'master' into REL1_36 - 13https://git.io/JOKA0 [14:01:00] Zppix: lol, don't remind me of how testwiki was dropped :'( [14:01:07] Ha [14:01:26] Zppix, yeah Reception123 and NDKilla have like no contributions on testwiki because of that [14:01:50] Where they not imported? [14:01:53] Were* [14:01:54] miraheze/mw-config - Reception123 the build passed. [14:02:00] miraheze/mw-config - Reception123 the build passed. [14:02:09] Zppix, yeah they were, but we'd have to run the fix imported edits script I think [14:02:17] to get their contribution totals to be updated [14:02:57] PROBLEM - intp.miraheze.org - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for intp.miraheze.org could not be found [14:04:31] RECOVERY - jobrunner3 Puppet on jobrunner3 is OK: OK: Puppet is currently enabled, last run 47 seconds ago with 0 failures [14:05:08] RECOVERY - jobrunner4 Puppet on jobrunner4 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [14:07:55] !log sudo -u www-data php /srv/mediawiki/w/maintenance/mergeMessageFileList.php --output /srv/mediawiki/config/ExtensionMessageFiles.php --wiki loginwiki --lang en on jbr*/mw* [14:08:00] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:08:08] PROBLEM - mw8 Current Load on mw8 is WARNING: WARNING - load average: 7.95, 6.88, 5.26 [14:08:09] !log without the --lang en part, that was a mistake [14:08:27] !log sudo -u www-data php /srv/mediawiki/w/maintenance/rebuildLocalisationCache.php --wiki loginwiki on mw*/jbr* [14:08:48] MirahezeLogbot: hello? [14:08:50] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:09:16] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:09:16] I am a logbot running on mon2. [14:09:16] Messages are logged to . [14:09:16] To log a message, type !log . [14:09:35] Reception123: MirahezeLogbot is out of the office can i take a message xD [14:09:50] RECOVERY - intp.miraheze.org - reverse DNS on sslhost is OK: rDNS OK - intp.miraheze.org reverse DNS resolves to cp11.miraheze.org [14:10:12] RECOVERY - mw8 Current Load on mw8 is OK: OK - load average: 5.45, 6.25, 5.22 [14:10:37] heh, that's now what we pay it to do! [14:10:51] PROBLEM - cloud4 Current Load on cloud4 is WARNING: WARNING - load average: 21.89, 18.21, 14.40 [14:10:55] MirahezeLogbot: a 30 second delay is simply unacceptable. Very disappointing behavior [14:10:56] I am a logbot running on mon2. [14:10:56] Messages are logged to . [14:10:56] To log a message, type !log . [14:11:36] PROBLEM - mw10 Current Load on mw10 is CRITICAL: CRITICAL - load average: 8.04, 6.84, 5.51 [14:12:03] Reception123, lol [14:12:11] at your comment to MirahezeLogbot [14:12:37] [02miraheze/mw-config] 07Reception123 pushed 031 commit to 03revert-3840-revert-3839-patch-161 [+0/-0/±2] 13https://git.io/JOPwy [14:12:38] [02miraheze/mw-config] 07Reception123 03c287ab3 - Revert "Revert "Convert SimpleTooltip to use ExtensionRegistry (#3839)" (#3840)" [14:12:40] [02mw-config] 07Reception123 created branch 03revert-3840-revert-3839-patch-161 - 13https://git.io/vbvb3 [14:12:45] [02mw-config] 07Reception123 opened pull request 03#3848: Convert SimpleTooltip to use ExtensionRegistry - 13https://git.io/JOPw9 [14:12:59] PROBLEM - cloud5 Current Load on cloud5 is WARNING: WARNING - load average: 20.93, 17.41, 13.98 [14:13:36] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 6.64, 6.70, 5.62 [14:13:55] miraheze/mw-config - Reception123 the build passed. [14:14:54] RECOVERY - cloud4 Current Load on cloud4 is OK: OK - load average: 16.76, 18.92, 15.68 [14:14:55] RECOVERY - cloud5 Current Load on cloud5 is OK: OK - load average: 17.81, 17.43, 14.40 [14:18:43] !log disabled puppet on mw* [14:18:47] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:18:50] [02mw-config] 07Reception123 closed pull request 03#3848: Convert SimpleTooltip to use ExtensionRegistry - 13https://git.io/JOPw9 [14:18:51] PROBLEM - cloud4 Current Load on cloud4 is WARNING: WARNING - load average: 20.21, 21.02, 17.38 [14:18:52] [02miraheze/mw-config] 07Reception123 pushed 031 commit to 03master [+0/-0/±2] 13https://git.io/JOPoI [14:18:53] [02miraheze/mw-config] 07Reception123 031d60dac - Revert "Revert "Convert SimpleTooltip to use ExtensionRegistry (#3839)" (#3840)" (#3848) [14:18:55] [02mw-config] 07Reception123 synchronize pull request 03#3845: Merge 'master' into REL1_36 - 13https://git.io/JOKA0 [14:18:56] [02miraheze/mediawiki] 07Reception123 pushed 031 commit to 03REL1_35 [+0/-0/±2] 13https://git.io/JOPot [14:18:58] [02miraheze/mediawiki] 07Reception123 03d1ee44b - Switch SimpleTooltip to Universal Omega fork (T7156) [14:20:00] miraheze/mw-config - Reception123 the build passed. [14:20:01] miraheze/mw-config - Reception123 the build passed. [14:20:34] PROBLEM - mw11 Puppet on mw11 is WARNING: WARNING: Puppet is currently disabled, message: Reception123 - fork switch, last run 13 minutes ago with 0 failures [14:20:40] dmehus: well test3 is already unhappy and is down [14:20:49] RECOVERY - cloud4 Current Load on cloud4 is OK: OK - load average: 12.64, 18.01, 16.73 [14:21:09] PROBLEM - mw10 Puppet on mw10 is WARNING: WARNING: Puppet is currently disabled, message: Reception123 - fork switch, last run 14 minutes ago with 0 failures [14:21:18] Reception123: whats error [14:21:20] if it's what I think it is (just takes a lot of time to complete) then I'll have to depool mw*s and do everything one by one [14:21:27] Zppix: oh it's a long story [14:21:39] Ok? [14:21:44] Zppix: it's basically because I'm switching from an extension with require_once to Omega's fork that uses wfLoad [14:21:50] PROBLEM - mw9 Puppet on mw9 is WARNING: WARNING: Puppet is currently disabled, message: Reception123 - fork switch, last run 14 minutes ago with 0 failures [14:22:00] PROBLEM - mw8 Puppet on mw8 is CRITICAL: CRITICAL: Failed to apply catalog, zero resources tracked by Puppet. It might be a dependency cycle. [14:22:02] and yup puppet just failed great, like last time [14:22:04] That should cause issues? [14:22:11] Shouldnt [14:22:15] hmm yeah, in theory [14:22:26] last time I tried to fix puppet failing Git errored too so I had to revert the whole thing [14:22:27] What ext is it [14:22:31] that's why I'm using this method now [14:22:34] SimpleTooltip [14:22:49] oh yes, it worked now! [14:23:22] (the fix for puppet) [14:24:17] ok then, guess I'll have to depool as the downtime is a bit too long 1-2 mins [14:25:32] !log depool mw8 and enable puppet on mw8 [14:25:35] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:26:24] Reception123, oh, looks like you got it fixed and test3wiki is working again [14:26:42] dmehus: yeah but there was a bit of downtime so it's safer (but more tedious) to depool mw and do it mw by mw [14:27:10] PROBLEM - cp10 Varnish Backends on cp10 is CRITICAL: 1 backends are down. mw8 [14:27:28] PROBLEM - cp12 Varnish Backends on cp12 is CRITICAL: 1 backends are down. mw8 [14:27:28] Reception123, ah why do you have to depool all the mw servers if it was just test3? [14:27:36] PROBLEM - cp3 Varnish Backends on cp3 is CRITICAL: 1 backends are down. mw8 [14:27:46] dmehus: yeah, one by one [14:27:52] so the change can go through and not actually cause downtime [14:28:01] Reception123, oh [14:28:08] because on test3 puppet failed at first and caused test3 to go down for a bit (like last time) [14:28:13] and to fix that I need to run something to fix it [14:28:27] PROBLEM - cp11 Varnish Backends on cp11 is CRITICAL: 1 backends are down. mw8 [14:28:29] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 9.35, 7.38, 6.04 [14:29:19] heh, the other mw*s aren't happy that mw8 is no longer doing their work [14:29:23] PROBLEM - mw10 Current Load on mw10 is CRITICAL: CRITICAL - load average: 10.93, 8.94, 6.98 [14:29:44] but yikes that's still very high for just one mw out [14:30:25] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 7.38, 7.36, 6.19 [14:31:19] Reception123, yeah, unless we have peak demand at the moment, perhaps? [14:31:41] dmehus: yeah, that's true too as this time is basically day in many timezones [14:31:53] US and Europe [14:31:55] or maybe the SimpleTooltip is an expensive parser function extension we can't afford? [14:32:00] and Asia it's close to night probably [14:32:06] yeah [14:32:13] oh no, it doesn't have anything to do with SimpleTooltip, the load is just beacause mw8 is idle [14:32:18] so the others have to take over [14:32:20] ah [14:32:21] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 9.72, 8.08, 6.58 [14:32:38] urgh, the fix failed again on mw8 [14:32:59] though now that it's depooled I can mess around with it without causing downtime at least [14:33:10] Reception123, oh I wonder why m28 is idle. Is there a way we can have traffic routed to under-utilized mw servers when the others are peaking? [14:33:24] dmehus: it's idle because I depooled it :) [14:33:26] > though now that it's depooled I can mess around with it without causing downtime at least [14:33:26] heh, true [14:33:28] so it's not doing anything at the moment [14:33:32] Reception123, oh right [14:33:45] oh so depool means to take out of service, basically [14:33:52] yeah, it can't serve requests anymore [14:33:56] ah [14:34:11] so it's up to the other three [14:34:40] * Reception123 doesn't like switching repos [14:35:09] oh [14:35:59] RhinosF1: hi [14:36:15] hey RhinosF1 [14:37:19] dmehus: see here https://grafana.miraheze.org/d/W9MIkA7iz/miraheze-cluster?viewPanel=281&orgId=1&var-job=node&var-node=mw8.miraheze.org&var-port=9100 [14:37:20] [ Grafana ] - grafana.miraheze.org [14:37:23] you can see the load going completely down [14:37:39] not at 0 though, as the server still has it's processes and it's also running puppet [14:38:24] Hi dmehus [14:38:26] RhinosF1: Notice: /Stage[main]/Mediawiki/Git::Clone[MediaWiki core]/Exec[git_pull_MediaWiki core]/returns: fatal: remote error: upload-pack: not our ref 5c515129a1e1184f0733e6a2a81cac18491141b3 [14:38:27] urgh [14:38:32] it didn't happen on test3 but it's happening on mw8 [14:38:32] PROBLEM - jobrunner3 Puppet on jobrunner3 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 2 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [14:38:34] and I don't get how to fix it [14:38:39] cd /srv/mediawiki/w/extensions && sudo -u www-data git reset --hard origin/REL1_35 && sudo -u www-data git pull ; sudo -u www-data git submodule update doesn't work [14:39:01] mw8 is depooled currently so what happened a few days ago didn't this time, as I knew something would go wrong again [14:39:07] Ok looking [14:39:09] PROBLEM - jobrunner4 Puppet on jobrunner4 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 3 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [14:39:20] Reception123: stop puppet everywhere [14:39:23] RhinosF1: I did [14:39:30] puppet is stopped on mw*, mw8 is depooled with puppet enabled [14:39:39] that was my plan from the start to avoid problems [14:39:40] Not jbr* then [14:39:53] I'll have a go [14:40:12] thanks, let me know what you find and feel free to do whatever on mw8 since it's out [14:40:14] [02mw-config] 07Universal-Omega commented on pull request 03#3846: add more file formats to wgFileExtensions - 13https://git.io/JOPi1 [14:40:21] and there's no point in trying test3 as it worked fine there [14:40:31] PROBLEM - jobrunner3 Puppet on jobrunner3 is WARNING: WARNING: Puppet is currently disabled, message: Reception123 - fork, last run 4 minutes ago with 1 failures [14:40:53] Reception123, ack, oof yeah [14:41:08] PROBLEM - jobrunner4 Puppet on jobrunner4 is WARNING: WARNING: Puppet is currently disabled, message: Reception123 - fork, last run 5 minutes ago with 1 failures [14:42:26] RhinosF1: Universal Omega proposed git reset --hard origin/master in /extensions/SimpleTooltip [14:42:32] should we try that? [14:42:58] Reception123: looking [14:44:15] great, let me know what happens [14:44:32] !log rhinos@mw8:/srv/mediawiki/w$ sudo -u www-data git submodule deinit extensions/SimpleTooltip -f [14:44:43] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:47:44] wow the other mw*s are really really going up [14:48:04] 13.09 at mw10's max [14:49:56] RhinosF1: any results? [14:51:07] What's up? [14:51:38] Reception123: think I fixed it [14:51:40] JohnLewis: basically Universal Omega created a fork for an unmaintained extension and in the process also switched from require_once to wfLoad [14:51:45] JohnLewis: hopefully nothing [14:51:52] Reception123: fixed [14:52:04] Reception123: Mkay, but why would that cause load of 13+? [14:52:06] and now I've switched the fork and because of a failure last time to be safe disabled puppet on mw* and depooled mw8 to first do it there [14:52:08] !log rhinos@mw8:/srv/mediawiki/w$ sudo -u www-data rm .git/modules/extensions/SimpleTooltip/config [14:52:14] RhinosF1: feel free to run puppet then [14:52:20] !log rhinos@mw8:/srv/mediawiki/w$ sudo -u www-data git submodule sync --recursive [14:52:20] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:52:33] !log rhinos@mw8:/srv/mediawiki/w$ cd /srv/mediawiki/w/extensions && sudo -u www-data git reset --hard origin/REL1_35 && sudo -u www-data git pull ; sudo -u www-data git submodule update --init [14:52:33] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:52:34] JohnLewis: because mw8 had to be disabled so that it wouldn't cause a 500 error (as it did last time) [14:52:48] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:52:51] RhinosF1: do you think we'd be able to do that on the other mws without having to deal with puppet failing and causing errors? [14:53:03] Reception123: changing an upstream url shouldn't cause a 500, it should be a simple CLI command to update it [14:53:26] JohnLewis: It seemed to be stuck on the old origin and wouldn't auto update [14:53:34] JohnLewis: we got Notice: /Stage[main]/Mediawiki/Git::Clone[MediaWiki core]/Exec[git_pull_MediaWiki core]/returns: fatal: remote error: upload-pack: not our ref 5c515129a1e1184f0733e6a2a81cac18491141b3 [14:53:35] it doesn't auto update [14:53:47] git submodule sync is how you change it if the config has changed [14:53:54] otherwise git submodule set-url [14:53:56] RECOVERY - mw8 Puppet on mw8 is OK: OK: Puppet is currently enabled, last run 37 seconds ago with 0 failures [14:54:08] JohnLewis: ye that didn't work until I deleted the old config [14:54:12] JohnLewis: oh, so you can use that to change URL for a submodule? [14:54:19] Reception123: yeah [14:54:25] !log repool mw8 [14:54:27] RECOVERY - cp11 Varnish Backends on cp11 is OK: All 7 backends are healthy [14:54:33] well that's good to know [14:54:44] !log sudo -u www-data php /srv/mediawiki/w/maintenance/mergeMessageFileList.php --output /srv/mediawiki/config/ExtensionMessageFiles.php --wiki loginwiki on mw8 [14:54:46] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:54:46] RECOVERY - cp10 Varnish Backends on cp10 is OK: All 7 backends are healthy [14:54:52] RhinosF1: I see you ran deinit, that would be why as you only only use sync on an init or entirely deinit'd submodule [14:54:55] !log sudo -u www-data php /srv/mediawiki/w/maintenance/rebuildLocalisationCache.php --wiki loginwiki on mw* [14:54:57] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:55:03] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [14:55:08] Reception123, do we have a documentation page we could add that note to? might be useful for the future [14:55:29] we could add it to updating extensions perhaps [14:55:30] RECOVERY - cp12 Varnish Backends on cp12 is OK: All 7 backends are healthy [14:55:30] RECOVERY - cp3 Varnish Backends on cp3 is OK: All 7 backends are healthy [14:55:35] JohnLewis: ah ok [14:55:39] Reception123, oh ok, cool :) [14:56:05] RhinosF1: do you think you could get it to work directly on mw9/mw10/mw11 without having puppet fail and having to depool? [14:56:08] if you can, use set-url before updating the repo if you want no puppet failures [14:56:38] Reception123: not really as I'm about to go into a careers event [14:56:56] oh [14:57:28] fyi https://phabricator.miraheze.org/T7179 isn't a CW issue [14:57:29] [ ⚓ T7179 CreateWiki AI fails to work due to memory issue ] - phabricator.miraheze.org [14:57:37] JohnLewis: so what would I have to do now that I've already done the other method (https://github.com/miraheze/mediawiki/commit/d1ee44ba7f6c1af84c4330113e5ddc434ac9c1a9) so that I don't need to depool the other servers? [14:57:38] [ Switch SimpleTooltip to Universal Omega fork (T7156) · miraheze/mediawiki@d1ee44b · GitHub ] - github.com [14:58:10] Reception123: I believe it's git submodule set-url extensions/ [14:58:26] ok, will try that then on mw9 [14:59:05] JohnLewis: is it a case of just upping the limit or? [14:59:19] PROBLEM - mw8 Current Load on mw8 is CRITICAL: CRITICAL - load average: 9.94, 6.05, 3.39 [14:59:28] RhinosF1: yes, because the file would have increased in size likely now [15:00:20] alternatively now I think about it... [15:00:35] Perhaps it's a case of there being insufficient memory on whichever server it ran on? [15:00:35] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 7.19, 7.03, 6.04 [15:01:26] hmm, doesn't seem to like reception@mw9:/srv/mediawiki/w/extensions$ sudo -u www-data git submodule set-url SimpleTooltip https://github.com/Universal-Omega/SimpleTooltip.git [15:01:27] [ GitHub - Universal-Omega/SimpleTooltip ] - github.com [15:01:34] Reception123: you just do sudo -u www-data git pull ; sudo -u www-data git submodule sync ; sudo -u www-data git submodule update [15:01:40] on each mw* and jobrunner* server [15:01:52] ok, let me try that then [15:02:05] Reception123: why didn't it like that? [15:02:40] it just gave the help instructions when I tried it [15:03:19] PROBLEM - mw8 Current Load on mw8 is WARNING: WARNING - load average: 7.95, 7.58, 4.69 [15:03:40] RECOVERY - mw9 Puppet on mw9 is OK: OK: Puppet is currently enabled, last run 2 minutes ago with 0 failures [15:03:42] !log sudo -u www-data git pull ; sudo -u www-data git submodule sync ; sudo -u www-data git submodule update on mw9 [15:03:47] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [15:04:05] [02mw-config] 07Universal-Omega opened pull request 03#3849: Update SimpleBlogPage config - 13https://git.io/JOPMj [15:04:28] PROBLEM - mw9 Current Load on mw9 is CRITICAL: CRITICAL - load average: 8.13, 7.70, 6.53 [15:04:48] Reception123: only guess would be because you didn't give the full path known which is extensions/SimpleTooltip [15:05:02] I also tried with extensions/SimpleTooltip [15:05:05] !log sudo -u www-data git pull ; sudo -u www-data git submodule sync ; sudo -u www-data git submodule update on mw10/mw11 [15:05:10] [02mw-config] 07Universal-Omega synchronize pull request 03#3849: Update SimpleBlogPage config - 13https://git.io/JOPMj [15:05:10] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [15:05:12] miraheze/mw-config - Universal-Omega the build passed. [15:05:21] Strange [15:05:39] !log sudo -u www-data git pull ; sudo -u www-data git submodule sync ; sudo -u www-data git submodule update on jbr3/jbr4 [15:05:42] [02mw-config] 07Universal-Omega synchronize pull request 03#3849: Update SimpleBlogPage config - 13https://git.io/JOPMj [15:05:44] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [15:05:45] yeah, at least Paladox's suggestion did the job [15:06:17] miraheze/mw-config - Universal-Omega the build passed. [15:06:25] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 6.32, 7.21, 6.49 [15:06:26] JohnLewis: regarding https://phabricator.miraheze.org/T7179 one thing didn't make sense. why is the tried to allocate number 327680 < Allowed memory size of 1572864000 ? [15:06:27] [ ⚓ T7179 CreateWiki AI fails to work due to memory issue ] - phabricator.miraheze.org [15:06:36] is it because the allowed memory is the total? [15:06:58] miraheze/mw-config - Universal-Omega the build passed. [15:07:02] RECOVERY - mw10 Puppet on mw10 is OK: OK: Puppet is currently enabled, last run 2 minutes ago with 0 failures [15:07:02] !log re-enabled puppet everywhere on mw* and jbr* [15:07:06] Logged the message at https://meta.miraheze.org/wiki/Tech:Server_admin_log [15:07:09] PROBLEM - jobrunner4 Puppet on jobrunner4 is CRITICAL: CRITICAL: Puppet has 1 failures. Last run 2 minutes ago with 1 failures. Failed resources (up to 3 shown): Exec[git_pull_MediaWiki core] [15:07:22] Reception123: ^ [15:07:43] still running so that must be from before [15:08:23] RECOVERY - mw9 Current Load on mw9 is OK: OK - load average: 5.21, 6.47, 6.30 [15:08:27] RECOVERY - mw11 Puppet on mw11 is OK: OK: Puppet is currently enabled, last run 17 seconds ago with 0 failures [15:08:31] RECOVERY - jobrunner3 Puppet on jobrunner3 is OK: OK: Puppet is currently enabled, last run 43 seconds ago with 0 failures [15:09:05] paladox: so next time (as there's a few more extensions to switch) would I just be fine directly running the command you gave me? [15:09:08] RECOVERY - jobrunner4 Puppet on jobrunner4 is OK: OK: Puppet is currently enabled, last run 1 minute ago with 0 failures [15:09:14] yes [15:09:19] RECOVERY - mw8 Current Load on mw8 is OK: OK - load average: 5.17, 6.73, 5.29 [15:10:16] Reception123: second comment I put about it potentially exceeding the total available memory on which ever server it ran on [15:10:44] ah [15:14:03] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 7.14, 7.20, 7.96 [15:14:46] On the load side though, guess the mw upgrade won't be fun then if depooling one mw server causes the others to reach 13 [15:14:55] Upgrading is never fun [15:15:06] dmehus: read up in here [15:15:14] yeah, it's going to be tedious [15:15:32] Specifically 16:06:26 JohnLewi.s: regarding https://phabricator.miraheze.org/T7179 one thing didn't make sense. why is the tried to allocate number 327680 < Allowed memory size of 1572864000 ? [15:15:33] [ ⚓ T7179 CreateWiki AI fails to work due to memory issue ] - phabricator.miraheze.org [15:15:43] RhinosF1: when you have time you'll also need to pull in these changes to REL1_36, though I need to finish with all the forks first [15:15:57] Reception123: I'm going to do it when rc0 comes out [15:16:15] Oh you mean make the changes you've done to 1.35 to 1.36 [15:16:22] Yeah please track them somehwere [15:16:33] yeah, I'll do it on the task [15:16:43] I also added an extension, so that will need to be done too [15:16:54] RhinosF1, yes [15:16:57] * RhinosF1 did send an email out saying all extension changes must be done on both branches [15:17:11] dmehus: that'll explain why your comment on phab isn't useful [15:17:23] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 4.99, 6.29, 7.72 [15:17:41] Yeah, I'll do it after I finish with them [15:18:04] Email said new exts must be added [15:18:23] But yeah please either sync or track it in the upgrade blockers column [15:19:42] RhinosF1: yeah, I'll sync after I'm done. Is there an easy way or do you think it's easier if I just redo the changes but on the REL1_36 branch? [15:19:56] (though I'll leave the rest of the work for tomorrow or this weekend) [15:20:20] Reception123: just redo on 1.36 branches [15:20:34] Because the history will have diverged and they'll be on seperate commits [15:20:39] In some cases [15:20:45] yeah, the other methods would be tricky [15:20:47] easier to just redo [15:21:48] Too much chance of it going weong [15:21:50] Wrong [15:23:56] RhinosF1, that would just apply to the commits to the /miraheze/mediawiki repo, right? /mw-config shouldn't be affected by the 1.35 and 1.36 switch, right? [15:24:02] RECOVERY - mw11 Current Load on mw11 is OK: OK - load average: 5.08, 5.35, 6.61 [15:25:21] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 5.70, 5.34, 6.69 [15:28:17] [02miraheze/puppet] 07JohnFLewis pushed 031 commit to 03master [+1/-0/±4] 13https://git.io/JOPH2 [15:28:18] [02miraheze/puppet] 07JohnFLewis 036330d1b - introduce jobrunner::intensive for high memory tasks [15:28:19] Reception123: ^ [15:29:31] thanks! [15:30:32] JohnLewis: where would that go when the Capacity Proposal goes through? [15:30:41] the "MediaWiki task server for heavy and intensive MediaWiki tasks "? [15:31:02] Yeah, seems best, and they are rare and irregular but high intensity tasks [15:31:28] yeah [15:32:02] JohnLewis, does https://git.io/JOPH2 allow jobrunner to run the CW AI jobs when SRE runs the script without them having to specify a minimum memory allocation? [15:32:03] [ Comparing d078b9bccbf9...6330d1b62500 · miraheze/puppet · GitHub ] - git.io [15:33:12] dmehus: the jobrunner always runs it with the allocation. Anyone running the script will always need to specify the allocation as well otherwise it won't be allocated [15:34:28] dmehus: mw-config will do but that's easier to handle [15:35:14] JohnLewis, oh okay, thanks. so that was the reason why the CreateWiki AI scores have not been updating, did we just need to implement this since forking jobrunner? [15:35:28] RhinosF1, ah okay, cool [15:35:40] It has nothing to do with forking jobrunner either [15:35:44] oh [15:36:01] I wonder why it hasn't been assigning scores since 17 April 2021 then [15:36:03] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 9.33, 7.41, 6.81 [15:36:11] PROBLEM - mw10 Current Load on mw10 is CRITICAL: CRITICAL - load average: 9.44, 7.11, 6.65 [15:36:13] as I would like to know the cause of that [15:36:17] Because jobs now run everywhere [15:36:24] Same reason icinga-miraheze won't shut up [15:36:54] But we're trying to avoid having to reimage just after creating a server and extra work at MediaWiki update time [15:36:56] > Because jobs now run everywhere [15:36:56] ah, can you elaborate on that a tad more? [15:37:01] Well no, if it hasn't since April 17th, it would be because of the model file being changed [15:37:11] JohnLewis, ah [15:37:17] and yeah because the AI file was updated [15:37:23] More data needs more memory [15:37:56] Which if that is the case, this won't have fixed the problem [15:38:03] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 7.09, 7.68, 7.01 [15:38:08] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 6.99, 7.13, 6.73 [15:38:15] JohnLewis, so do you mean it's not automatically run then? It has to be run periodically by SRE, until it's next changed or we want to update the scores? [15:38:20] also the load on mw* has been a problem for a lot longer [15:38:41] PROBLEM - m.miraheze.org - LetsEncrypt on sslhost is CRITICAL: CRITICAL - Certificate 'm.miraheze.org' expires in 7 day(s) (Fri 30 Apr 2021 15:38:35 GMT +0000). [15:38:47] dmehus: the fix is probably less effort than running it manually [15:40:04] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 8.21, 8.41, 7.39 [15:40:25] JohnLewis, ah, okay...I'm still a bit confused on the technicals of how it works. Perhaps it'd help if you or someone could point me to the location of the model file? [15:42:03] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 5.63, 6.64, 6.63 [15:42:04] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 6.39, 7.85, 7.32 [15:42:13] dmehus: it's on server [15:42:20] dmehus: it probably wouldn't, but https://static.miraheze.org/requestmodel.phpml [15:42:27] oh [15:42:49] But simply put updating the model that CW AI uses to have more data makes it take more memory to process it because it's bigger [15:43:41] RhinosF1, yeah, that's true, but did we implement any changes recently since April 17th that would've changed that? [15:44:18] [02puppet] 07Universal-Omega opened pull request 03#1744: Remove composer for Validator - 13https://git.io/JOP5S [15:44:19] I can't see any changes to the RequestWikiAIJob.php, the most recent one was 14 March 2021, and we've had AI scores since then [15:44:36] [02mw-config] 07Universal-Omega opened pull request 03#3850: Remove Validator - 13https://git.io/JOP5Q [15:45:18] dmehus: https://meta.miraheze.org/w/index.php?title=Tech:Server_admin_log&diff=171707&oldid=171692&diffmode=source essentially [15:45:20] [ Difference between revisions of "Tech:Server admin log" - Miraheze Meta ] - meta.miraheze.org [15:45:43] miraheze/mw-config - Universal-Omega the build passed. [15:46:02] JohnLewis, ah [15:46:02] RECOVERY - mw11 Current Load on mw11 is OK: OK - load average: 4.71, 6.28, 6.80 [15:46:50] so for whatever reason, re-running the script should fix it I guess? [15:47:11] No, as re-running it would generate it the same way [15:49:47] oh [15:59:46] PROBLEM - wiki.seredrau.xyz - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for wiki.seredrau.xyz could not be found [15:59:48] PROBLEM - wiki.aridia.space - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for wiki.aridia.space could not be found [15:59:48] PROBLEM - stablestate.org - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for stablestate.org could not be found [15:59:49] PROBLEM - idleon.info - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for idleon.info could not be found [15:59:53] PROBLEM - wiki.hrznstudio.com - reverse DNS on sslhost is WARNING: rDNS WARNING - reverse DNS entry for wiki.hrznstudio.com could not be found [16:04:53] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 7.14, 6.53, 5.57 [16:05:19] PROBLEM - mw8 Current Load on mw8 is WARNING: WARNING - load average: 7.49, 6.59, 5.74 [16:06:39] RECOVERY - wiki.aridia.space - reverse DNS on sslhost is OK: rDNS OK - wiki.aridia.space reverse DNS resolves to cp10.miraheze.org [16:06:40] RECOVERY - stablestate.org - reverse DNS on sslhost is OK: rDNS OK - stablestate.org reverse DNS resolves to cp10.miraheze.org [16:06:40] RECOVERY - idleon.info - reverse DNS on sslhost is OK: rDNS OK - idleon.info reverse DNS resolves to cp10.miraheze.org [16:06:40] RECOVERY - wiki.seredrau.xyz - reverse DNS on sslhost is OK: rDNS OK - wiki.seredrau.xyz reverse DNS resolves to cp11.miraheze.org [16:06:49] RECOVERY - mw9 Current Load on mw9 is OK: OK - load average: 5.16, 6.03, 5.50 [16:06:55] RECOVERY - wiki.hrznstudio.com - reverse DNS on sslhost is OK: rDNS OK - wiki.hrznstudio.com reverse DNS resolves to cp10.miraheze.org [16:07:19] RECOVERY - mw8 Current Load on mw8 is OK: OK - load average: 5.67, 6.34, 5.76 [16:08:28] PROBLEM - cp11 Current Load on cp11 is CRITICAL: CRITICAL - load average: 3.77, 4.35, 2.34 [16:10:28] RECOVERY - cp11 Current Load on cp11 is OK: OK - load average: 1.60, 3.34, 2.20 [16:22:36] Reception123: https://phabricator.miraheze.org/T7173#142475 [16:22:37] [ ⚓ T7173 15m load average on mw* has been steadily rising since 6am ] - phabricator.miraheze.org [16:23:38] JohnLewis: will close then, not much we can do except implement the capacity plan [16:33:05] Reception123: and if https://phabricator.miraheze.org/T7179 isn't fixed by my change above, it'll require someone to change the memory limit in the jobrunner config file to something appropriate [16:33:06] [ ⚓ T7179 CreateWiki AI fails to work due to memory issue ] - phabricator.miraheze.org [16:34:36] JohnLewis: if re-running the script won't fix the issue, how would we try to fix before having to change the memory limit? And if that doesn't work, what limit would you say would be appropriate to try out? [16:35:18] You can't fix it without not changing the memory limit because the memory limit would be the problem :) [16:35:46] JohnLewis: oh, but you said if it isn't fixed it will require someone to change the memory limit [16:35:51] so I thought you meant your change alone would fix [16:36:15] *might fix, and only if it doesn't a change would have to be made [16:36:31] Well, if the problem is there's a lack of memory on the server its running on, it would because jobrunner3 has like 90% free memory or something [16:36:43] ah [16:37:10] so then we'll have to change the memory limit anyway, though not sure what we could up it to [16:37:52] Has it failed since my change above was deployed? [16:39:14] doesn't look like it [16:41:37] [02miraheze/puppet] 07JohnFLewis pushed 031 commit to 03master [+0/-0/±1] 13https://git.io/JOXeN [16:41:39] [02miraheze/puppet] 07JohnFLewis 03b060d40 - fix hiera key [16:41:49] impossible to say because there was a typo in the hiera key :) [16:43:16] oh heh [16:51:40] Reception123: failed again, so indeed it is lack of allocated memory [16:54:11] hmm ok, what would you suggest we change the limit to? [16:56:00] If I knew, I'd do it now. Someone needs to look at what an appropriate limit it - once where the script runs but doesn't use more memory than necessary realistically [16:59:33] Ok [17:10:11] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 7.22, 6.93, 6.28 [17:12:08] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 6.02, 6.43, 6.16 [17:14:02] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 7.90, 6.94, 6.22 [17:16:03] RECOVERY - mw11 Current Load on mw11 is OK: OK - load average: 6.24, 6.61, 6.19 [17:35:19] PROBLEM - mw8 Current Load on mw8 is WARNING: WARNING - load average: 7.02, 6.46, 5.54 [17:37:19] RECOVERY - mw8 Current Load on mw8 is OK: OK - load average: 6.04, 6.45, 5.66 [18:00:32] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 9.98, 8.45, 7.20 [18:05:04] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 7.29, 6.81, 5.98 [18:05:20] PROBLEM - mw8 Current Load on mw8 is CRITICAL: CRITICAL - load average: 8.32, 7.33, 6.30 [18:07:02] PROBLEM - mw9 Current Load on mw9 is CRITICAL: CRITICAL - load average: 10.62, 8.09, 6.54 [18:09:20] PROBLEM - mw8 Current Load on mw8 is WARNING: WARNING - load average: 5.82, 7.25, 6.55 [18:13:54] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 6.67, 7.84, 7.80 [18:19:20] PROBLEM - mw8 Current Load on mw8 is CRITICAL: CRITICAL - load average: 8.08, 7.36, 6.92 [18:20:03] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 5.23, 6.40, 7.67 [18:21:19] RECOVERY - mw8 Current Load on mw8 is OK: OK - load average: 5.22, 6.50, 6.65 [18:21:42] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 3.46, 5.17, 6.60 [18:22:37] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 5.09, 7.51, 7.47 [18:26:33] RECOVERY - mw9 Current Load on mw9 is OK: OK - load average: 4.06, 5.61, 6.69 [18:32:02] RECOVERY - mw11 Current Load on mw11 is OK: OK - load average: 5.21, 5.87, 6.67 [18:54:00] PROBLEM - mw10 Current Load on mw10 is CRITICAL: CRITICAL - load average: 11.07, 7.85, 6.60 [18:55:58] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 7.51, 7.95, 6.81 [18:57:56] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 4.32, 6.56, 6.43 [19:36:22] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 7.23, 7.16, 6.70 [19:38:21] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 5.84, 6.69, 6.58 [19:46:07] PROBLEM - mw10 Current Load on mw10 is WARNING: WARNING - load average: 7.00, 7.54, 6.98 [19:48:04] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 5.12, 6.67, 6.73 [20:03:41] PROBLEM - wiki.mlpwiki.net - reverse DNS on sslhost is CRITICAL: rDNS CRITICAL - wiki.mlpwiki.net reverse DNS resolves to 192-185-16-85.unifiedlayer.com [20:05:16] PROBLEM - mw11 Current Load on mw11 is CRITICAL: CRITICAL - load average: 8.94, 6.80, 6.06 [20:05:39] PROBLEM - mw10 Current Load on mw10 is CRITICAL: CRITICAL - load average: 8.52, 6.91, 6.38 [20:05:47] PROBLEM - mw9 Current Load on mw9 is WARNING: WARNING - load average: 7.84, 6.67, 5.67 [20:07:12] PROBLEM - mw11 Current Load on mw11 is WARNING: WARNING - load average: 7.49, 6.98, 6.22 [20:07:36] RECOVERY - mw10 Current Load on mw10 is OK: OK - load average: 6.18, 6.72, 6.38 [20:07:46] RECOVERY - mw9 Current Load on mw9 is OK: OK - load average: 4.13, 5.82, 5.49 [20:09:08] RECOVERY - mw11 Current Load on mw11 is OK: OK - load average: 4.60, 5.84, 5.88 [21:29:03] .uptime [21:29:03] I've been sitting here for 5 days, 8:00:20 and I keep going! [21:29:08] .uptime dmehus [21:29:08] I've been sitting here for 5 days, 8:00:25 and I keep going! [21:29:18] hrm [21:29:21] oh right [21:29:59] I've been sitting here for online for 8 days (idle for less than a minute), MirahezeBot :P [21:31:39] oh wow, robla wins the uptime award of users in this channel (16 days), with SPF|Cloud not far behind at 13 days [21:33:00] (excluding Not-818c, at 28 days) [22:28:29] PROBLEM - cp12 Current Load on cp12 is WARNING: WARNING - load average: 1.98, 1.58, 1.23 [22:30:26] RECOVERY - cp12 Current Load on cp12 is OK: OK - load average: 1.33, 1.40, 1.20 [22:38:18] PROBLEM - cp12 Current Load on cp12 is CRITICAL: CRITICAL - load average: 1.56, 2.06, 1.58 [22:40:17] PROBLEM - cp12 Current Load on cp12 is WARNING: WARNING - load average: 1.52, 1.83, 1.55 [22:42:13] RECOVERY - cp12 Current Load on cp12 is OK: OK - load average: 1.39, 1.63, 1.50 [23:03:40] RECOVERY - wiki.mlpwiki.net - reverse DNS on sslhost is OK: rDNS OK - wiki.mlpwiki.net reverse DNS resolves to cp11.miraheze.org [23:10:57] Interesting.... [23:12:41] PROBLEM - wiki.mlpwiki.net - reverse DNS on sslhost is CRITICAL: rDNS CRITICAL - wiki.mlpwiki.net reverse DNS resolves to 192-185-16-85.unifiedlayer.com