[00:08:26] (03PS1) 10EBernhardson: Edit Project Config [extensions/Wikibase] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/485978 [00:09:54] (03Abandoned) 10EBernhardson: Edit Project Config [extensions/Wikibase] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/485978 (owner: 10EBernhardson) [00:17:11] (03PS1) 10Thcipriani: jjb: docker tox image update for apt security [integration/config] - 10https://gerrit.wikimedia.org/r/485983 [00:26:23] (03CR) 10Thcipriani: [C: 03+2] "deployed" [integration/config] - 10https://gerrit.wikimedia.org/r/485983 (owner: 10Thcipriani) [00:29:20] (03Merged) 10jenkins-bot: jjb: docker tox image update for apt security [integration/config] - 10https://gerrit.wikimedia.org/r/485983 (owner: 10Thcipriani) [00:36:18] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10dbarratt) Was the wmf.14 branch cut? [00:54:12] Krinkle: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/VisualEditor/+/485328 failing suggests that either ResourcesTest::testMissingDependencies is bypassing your new system (how?) or it's evaluated too early? [00:55:16] James_F: Haven't rule out either yet, but theres definitely another issue at the moment (see inline commenet) [00:55:30] Krinkle: Oh, the name is now computed? [00:55:35] Funky. [00:56:19] James_F: Yeah, I want to be strong in both 1) encouraging very hard to only use one module (there is no reason not to), and 2) leave the door open to this not being a module. We're just registering tests to execute and actual prod modules to load first. [00:56:27] Could potentially become like core, where all test are just 1 module. [00:56:28] * James_F nods. [00:57:03] like both core and ext combined just 1 bundle. It's all in the same req anyway, but just removes more indirection we don't need and things to think about for developers :) [00:57:08] anyway, that's for another day maybe [01:04:01] Krinkle: (Same error without the level.) [01:04:52] I can't update puppet on deployment-prep, there are merge conflicts with existing cherry picked changes [01:27:00] legoktm says it's meant to alert via shinken when that happens, but I don't know how to check if it has alerted [01:27:43] I logged in to deployment-puppetmaster03, went to /var/lib/git/operations/puppet , did "git rebase origin/production" [01:27:49] and it failed at [01:27:51] Applying: varnish: move $all_networks to $trusted_networks [01:28:10] with conflicts in three files [01:31:12] hrm, sometimes you have to fight with that repo a bit. It might be easy to fix if someone cherry picked something and forgot to remove the cherry-pick once it's merged. [01:31:59] the change was by Alexandros Kosiaris I4f6a6684d4337e3e7e670a2333d4e702ff21f5be [01:32:19] and it looks like it merged [01:32:26] lemme try to remove that one [01:33:13] that seemed to have done the trick [01:33:55] thanks, so it is up to date now? [01:34:25] yep, looks to be 15 commits ahead of master (should be changes in gerrit tagged "beta-picked") [01:34:36] er, ahead of "production" [02:17:21] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) @dbarratt: yes, it was logged above, see: >>! In T206668#4900697, @Stashbot wrote: > {nav icon=file, name=Mentioned in SAL (#wi... [02:17:51] obviously there is no alert [02:18:18] in git-sync-upstream.log I now see "error: could not apply c7262be... LOCAL HACK: Add puppetdb passwords" [02:20:01] at least my LocalSettings.php change is deployed [02:24:42] TimStarling: there used to be this shriken - Nov 17 04:09:32 PROBLEM - Long lived cherry-picks on puppetmaster on deployment-puppetmaster03 is CRITICAL: CRITICAL: 100.00% of data above the critical threshold [0.0] [02:25:24] Could the upgrade to stretch have removed it? [02:25:55] I couldn't even ssh to shinken [02:25:58] $ ssh shinken-02.shinken.eqiad.wmflabs [02:25:58] Connection closed by UNKNOWN port 65535 [02:26:34] Wrong one [02:26:36] I think [02:26:53] Maybe I’m wrong, though it was moved to stretch l [02:26:54] it says at https://tools.wmflabs.org/openstack-browser/project/shinken that that instance is responsible for shinken.wmflabs.org [02:27:11] that is the right one [02:27:41] you're just not in the group [02:28:53] apparently /var/log/ircecho/irc-releng.log was last modified on the 9th of December [02:29:34] Perm issues? [03:26:26] 10Release-Engineering-Team: shinken-wm hasn't alerted in -releng since 2018-12-09 - https://phabricator.wikimedia.org/T214455 (10Peachey88) [03:26:49] 10Release-Engineering-Team: shinken-wm hasn't alerted in -releng since 2018-12-09 - https://phabricator.wikimedia.org/T214455 (10Peachey88) [03:49:23] why is there no form of scap that pushes out a list of files, instead of a single file? [03:50:12] I mean, I know it was my fault in the first place, it was due to laziness, for the record, but I would have thought at some point the decision would have been revisited [04:40:38] 10Beta-Cluster-Infrastructure, 10Release-Engineering-Team, 10Shinken, 10monitoring: shinken-wm hasn't alerted in -releng since 2018-12-09 - https://phabricator.wikimedia.org/T214455 (10greg) [06:14:10] (03PS1) 10Rauppath: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 [06:14:14] (03CR) 10Welcome, new contributor!: "Thank you for making your first contribution to Wikimedia! :) To learn how to get your code changes reviewed faster and more likely to get" [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [06:28:43] (03Abandoned) 10Rauppath: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [06:29:28] (03CR) 10Rauppath: "na" [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [06:39:26] (03Restored) 10Rauppath: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [07:17:03] 10Release-Engineering-Team (Backlog), 10Release Pipeline, 10serviceops: Pipeline: provide a way to rebuild all blubber images - https://phabricator.wikimedia.org/T214431 (10greg) [07:21:53] 10Release-Engineering-Team (Kanban), 10MediaWiki-extensions-MultimediaViewer, 10Multimedia, 10User-zeljkofilipin: Misconfigured -- Unsupported OS/browser/version/device combo - https://phabricator.wikimedia.org/T214389 (10greg) [07:24:09] 10Gerrit, 10Release-Engineering-Team (Backlog): Upgrade to Gerrit 2.15.9 - https://phabricator.wikimedia.org/T214359 (10greg) [07:24:20] 10Gerrit, 10Release-Engineering-Team (Backlog): Upgrade to Gerrit 2.15.9 - https://phabricator.wikimedia.org/T214359 (10greg) p:05Triage→03Normal [07:37:20] (03CR) 10Rauppath: [C: 03+1] adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [07:41:07] (03Abandoned) 10Legoktm: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [08:16:15] 10Release-Engineering-Team (Backlog), 10Release Pipeline, 10serviceops: Pipeline: provide a way to rebuild all blubber images - https://phabricator.wikimedia.org/T214431 (10Joe) The general idea (specified in some ticket I lost track of, to be added later) for long-term support is: - when an image is ready... [08:25:20] (03CR) 10Thiemo Kreuz (WMDE): [C: 03+1] "Totally agree." (031 comment) [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/485732 (owner: 10Tim Starling) [08:39:05] 10Release-Engineering-Team (Kanban), 10docker-pkg, 10serviceops, 10Patch-For-Review: Some HEAD requests to docker-registry yields 405 Method not allowed - https://phabricator.wikimedia.org/T214441 (10Joe) FWIW, https://docker-registry.wikimedia.org/v2/releng/composer-php73/manifests/0.1.4 now responds with... [08:57:28] (03Restored) 10Rauppath: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [08:57:35] 10Release-Engineering-Team (Kanban), 10docker-pkg, 10serviceops, 10Patch-For-Review: Some HEAD requests to docker-registry yields 405 Method not allowed - https://phabricator.wikimedia.org/T214441 (10Joe) The bug is in the docker registry itself. It returns consistently 405 to HEAD requests to that URL: `... [08:58:34] 10Release-Engineering-Team (Kanban), 10docker-pkg, 10serviceops, 10Patch-For-Review: Some HEAD requests to docker-registry yields 405 Method not allowed - https://phabricator.wikimedia.org/T214441 (10Joe) On second thoughts - we should probably use a GET in docker-pkg since that's more reliable. [09:06:19] (03PS1) 10Thiemo Kreuz (WMDE): Fix double not sniff for (rare) edge-cases with instanceof [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) [09:07:42] (03Abandoned) 10Rauppath: adding new line in one of the class [pywikibot/core] (refs/meta/config) - 10https://gerrit.wikimedia.org/r/486006 (owner: 10Rauppath) [09:08:56] (03CR) 10Thiemo Kreuz (WMDE): "I tried to think of a few worst-case scenarios for this sniff:" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/483946 (https://phabricator.wikimedia.org/T203800) (owner: 10Mainframe98) [09:16:03] (03CR) 10Mainframe98: "> Patch Set 5: Code-Review+1" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/483946 (https://phabricator.wikimedia.org/T203800) (owner: 10Mainframe98) [09:28:28] 10Gerrit, 10Release-Engineering-Team (Backlog): Upgrade to Gerrit 2.15.9 - https://phabricator.wikimedia.org/T214359 (10Paladox) [09:28:31] 10Release-Engineering-Team (Kanban), 10Education-Program-Dashboard, 10MediaWiki-extensions-EducationProgram, 10Epic, 10User-greg: Deprecate and remove the EducationProgram extension from Wikimedia servers after June 30, 2018 - https://phabricator.wikimedia.org/T125618 (10Aklapper) [09:36:49] zeljkof: o/ Morning! I'm wondering have you ever had problems with browsertests failing as MWBot fails to login with WrongToken? [09:37:05] tarrow: morning! [09:37:11] hm, I don't think so [09:37:28] It suddenly happened to me yesterday; I asked around at WMDE and other people have had it too [09:37:43] looks like there is a few wrongtoken tasks in phab WrongToken [09:37:44] this is running them locally [09:37:52] sorry, https://phabricator.wikimedia.org/search/query/GI4fcKYGmpsZ/#R [09:37:59] hm [09:38:49] tarrow: could you please create a phab task, so we can track it there? [09:39:14] maybe something changed in mwbot or mw api? [09:39:23] zeljkof: sure! I'm 100% certain it is a local problem for me [09:39:49] but clearly other people have somehow go tin the same situation too [09:40:07] we can always close the task if it's only local [09:40:29] from a quick look, I don't think other WrongToken related phab tasks are, well, related [09:40:32] but I might be wrong [09:40:37] (03CR) 10Mainframe98: "This edge-case is headache inducing and scary. Did you find this in code somewhere?" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [10:41:07] 10Release-Engineering-Team (Kanban), 10MediaWiki-extensions-MultimediaViewer, 10Multimedia, 10User-zeljkofilipin: Misconfigured -- Unsupported OS/browser/version/device combo - https://phabricator.wikimedia.org/T214389 (10zeljkofilipin) According to [[ https://gerrit.wikimedia.org/r/plugins/gitiles/integra... [10:45:38] zeljkof: https://phabricator.wikimedia.org/T214471 [10:46:58] tarrow: just saw it in mail, I've added a few tags so I don't forget about it, but next week is All Hands in San Francisco, I will not have the time to look at this this week since I'm finishing some urgent things, and I'll be traveling all next week [10:47:28] I'll take a look (if nobody does until then) when I recover from jet lag [10:48:40] 10Release-Engineering-Team (Kanban), 10MediaWiki-Core-Tests, 10Browser-Tests, 10User-zeljkofilipin: wdio browser tests fail locally with Could not login: WrongToken and NeedToken - https://phabricator.wikimedia.org/T214471 (10zeljkofilipin) [10:54:02] 10Project-Admins: Convenient Discussions project - https://phabricator.wikimedia.org/T214291 (10Iniquity) @Aklapper we need you :) Is it necessary to create gerrit repo or git repo is enough? [11:20:24] zeljkof: thanks! I've "worked around" the problem by making someone else run them for now so no hurry. Take it easy :) [12:14:14] (03CR) 10Thiemo Kreuz (WMDE): "> Did you find this in code somewhere?" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [13:29:28] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [13:32:36] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [13:33:11] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [13:36:52] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [13:37:16] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [13:39:10] 10Continuous-Integration-Config, 10Operations, 10Puppet: puppet.git rake fails with ruby 2.5 - https://phabricator.wikimedia.org/T208566 (10fgiunchedi) In case it is useful, on a buster system I'm using this to run `rake` locally: `PUPPET_GEM_VERSION=4.10.12 bundle exec rake test` [13:55:32] 10Continuous-Integration-Config, 10Operations, 10Puppet: puppet.git rake fails with ruby 2.5 - https://phabricator.wikimedia.org/T208566 (10jbond) I have also been trying to get this working the following is also a useful data point https://bugzilla.redhat.com/show_bug.cgi?id=1440710, however applying that f... [14:30:37] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [14:36:32] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [14:52:09] (03PS5) 10Hashar: docker: add missing control files [integration/config] - 10https://gerrit.wikimedia.org/r/485199 [14:52:11] (03PS4) 10Hashar: (DO NOT SUBMIT) inspect docker deps [integration/config] - 10https://gerrit.wikimedia.org/r/485067 [14:54:29] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [15:01:19] 10Release-Engineering-Team (Kanban), 10User-zeljkofilipin: The first Selenium test for AbuseFilter - https://phabricator.wikimedia.org/T214478 (10zeljkofilipin) p:05Triage→03Normal [15:02:46] 10Release-Engineering-Team (Kanban), 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [15:04:45] 10Release-Engineering-Team (Kanban), 10AbuseFilter, 10Patch-For-Review, 10User-zeljkofilipin: The first Selenium test for AbuseFilter - https://phabricator.wikimedia.org/T214478 (10Daimona) [15:04:47] 10Release-Engineering-Team (Kanban), 10AbuseFilter, 10Patch-For-Review, 10User-zeljkofilipin: The first Selenium test for AbuseFilter - https://phabricator.wikimedia.org/T214478 (10zeljkofilipin) a:05zeljkofilipin→03Daimona [15:05:10] 10Release-Engineering-Team (Kanban), 10AbuseFilter, 10Patch-For-Review, 10User-zeljkofilipin: The first Selenium test for AbuseFilter - https://phabricator.wikimedia.org/T214478 (10zeljkofilipin) Assigned to @Daimona since he's already working on this. [15:05:45] 10Release-Engineering-Team (Kanban), 10AbuseFilter, 10Patch-For-Review, 10User-zeljkofilipin: The first Selenium test for AbuseFilter - https://phabricator.wikimedia.org/T214478 (10zeljkofilipin) [15:07:50] 10Release-Engineering-Team (Kanban), 10TimedMediaHandler, 10User-zeljkofilipin: The first Selenium test for TimedMediaHandler - https://phabricator.wikimedia.org/T214480 (10zeljkofilipin) p:05Triage→03Normal [15:09:16] 10Release-Engineering-Team (Kanban), 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [15:17:03] 10Release-Engineering-Team (Kanban), 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [15:18:09] 10Release-Engineering-Team (Kanban), 10User-zeljkofilipin: 5 of the 15 prioritized repositories have at least 1 end-to-end test - https://phabricator.wikimedia.org/T206621 (10zeljkofilipin) [15:34:42] 10Continuous-Integration-Config, 10Operations, 10Puppet: puppet.git rake fails with ruby 2.5 - https://phabricator.wikimedia.org/T208566 (10jbond) > https://bugzilla.redhat.com/show_bug.cgi?id=1440710, however applying that fix just causes a different error in fact, the patch mentioned in the commit above i... [15:46:44] 10Release-Engineering-Team (Kanban), 10MediaWiki-extensions-MultimediaViewer, 10Multimedia, 10User-zeljkofilipin: Misconfigured -- Unsupported OS/browser/version/device combo - https://phabricator.wikimedia.org/T214389 (10Jdlrobson) >>! In T214389#4901733, @zeljkofilipin wrote: > According to [[ https://ge... [16:19:52] 10Release-Engineering-Team (Kanban), 10MediaWiki-extensions-MultimediaViewer, 10Multimedia, 10User-zeljkofilipin: Misconfigured -- Unsupported OS/browser/version/device combo - https://phabricator.wikimedia.org/T214389 (10MarkTraceur) Looks like @zeljkofilipin wrote the config for the browser tests in Mult... [16:25:41] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.32.0-wmf.10 deployment blockers - https://phabricator.wikimedia.org/T191056 (10greg) [16:25:45] 10Release-Engineering-Team (Kanban), 10MediaWiki-Maintenance-scripts, 10MediaWiki-Vendor, 10Performance-Team (Radar), 10User-greg: Running scap sync-dir php-1.32.0-wmf.10 fails due to syntax error - https://phabricator.wikimedia.org/T198422 (10greg) 05Open→03Resolved a:03greg >>! In T198422#4639611... [17:12:29] legoktm: _joe_ thcipriani : some experiment to trace CI docker container dependencies https://phabricator.wikimedia.org/P8028 ;D [17:12:36] gotta copy paste a bit then you get some graph of the deps [17:19:39] 10Gerrit, 10Release-Engineering-Team: Delete es6 branch of Wikibase - https://phabricator.wikimedia.org/T214496 (10EBernhardson) [17:19:54] ^ should be very easy, but i don't know who would have the rights / who i should ask [17:24:37] 10Gerrit, 10Release-Engineering-Team: Delete es6 branch of Wikibase - https://phabricator.wikimedia.org/T214496 (10Lucas_Werkmeister_WMDE) 05Open→03Resolved a:03Lucas_Werkmeister_WMDE done, couldn’t do it via `git push` but was allowed to do it on https://gerrit.wikimedia.org/r/admin/projects/mediawiki/e... [17:29:18] 10Continuous-Integration-Config, 10Release-Engineering-Team (Backlog): Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10cooltey) [17:56:11] thcipriani: so yeah npm I don't know [17:56:16] well actually [17:56:32] One just need to s/python-minimal/python2.7/ [17:56:36] which would fix node-gyp [17:56:52] https://phabricator.wikimedia.org/T213944 [17:57:05] that's in the base releng/npm image? [17:57:07] so probably we should just do that as part of the rebuild [17:57:13] in two or three images [17:57:30] I'll grep as part of update [17:57:39] I looked at it late last week [17:57:42] https://etherpad.wikimedia.org/p/ci-apt-docker-why [17:57:48] eventually ended up with A HUGE YAK TO SHAVE [17:57:51] was the etherpad we used for updates yesterday [17:58:00] which has lead me to finding the docker registry yielding HTTP 405 Method not Allowed [17:58:24] $ git grep -l python-minimal [17:58:24] dockerfiles/java8-xgboost/changelog [17:58:24] dockerfiles/node10-test/Dockerfile.template [17:58:24] dockerfiles/npm-stretch/Dockerfile.template [17:58:24] dockerfiles/npm-test/changelog [17:58:24] dockerfiles/npm/Dockerfile.template [17:58:26] dockerfiles/npm6/Dockerfile.template [17:58:43] the line got copy pasted between the containers definition [17:59:08] which also got me to waste half a day to try to figure out a way to compose snippets of Dockerfiles using Jinja templates [17:59:12] but that went nowhere :( [18:00:20] so in short for npm/node ones just s/python-minimal/python2.7/ then dch all the containers that descend from them [18:01:06] k, I'll give it a shot [18:01:27] and that has lead me to https://phabricator.wikimedia.org/P8028 [18:02:01] determine how the containers are depending on each others based on dockerfiles/*/control files (and Depends: field) [18:02:04] which got me edges [18:02:11] and I let Graphviz to figur eit out [18:02:59] (03CR) 10Hashar: "example output https://phabricator.wikimedia.org/P8028" [integration/config] - 10https://gerrit.wikimedia.org/r/485067 (owner: 10Hashar) [18:04:09] I gotta finish preparing dinnre [18:41:23] (03CR) 10Mainframe98: [C: 04-1] "> Patch Set 1:" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [18:51:57] gitiles is now a core gerrit plugin! [19:00:00] Phabricator Upgrade Planning meeting time [19:00:30] (03CR) 10Mainframe98: [C: 04-1] "Really, the sane thing to do would be bailing after spotting T_INSTANCEOF, since it's only going to cause trouble. The warning message in " [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [19:04:01] (03CR) 10Mainframe98: [C: 04-1] "> I've attempted to do this, and it's quite trivial to realize, so I'm going ahead and integrate it in 483946, as it's low hanging fruit." [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [19:21:55] oh, we can create a plugin for site notice in PolyGerrit. [19:39:20] * paladox is working on a brand new zuul status plugin for gerrit exclusively for PolyGerrit's ui! [19:45:42] 10Phabricator, 10Release-Engineering-Team (Backlog): Implement phabricator database clustering support - https://phabricator.wikimedia.org/T112776 (10mmodell) [19:45:45] 10Phabricator, 10Release-Engineering-Team (Kanban), 10Availability, 10Patch-For-Review, and 2 others: Deploy phabricator to phab2001.codfw.wmnet - https://phabricator.wikimedia.org/T137928 (10mmodell) [19:48:01] (03CR) 10Thiemo Kreuz (WMDE): "> It's a hypothetical then?" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [19:56:21] 10Phabricator, 10Release-Engineering-Team (Kanban), 10DBA, 10Documentation, 10User-MModell: Prepare a disaster recovery plan for failing over from phab1001 to phab2001 (or phab2001 to 1001) - https://phabricator.wikimedia.org/T190572 (10mmodell) [19:56:25] 10Phabricator, 10Release-Engineering-Team (Kanban), 10Availability, 10Patch-For-Review, and 2 others: Deploy phabricator to phab2001.codfw.wmnet - https://phabricator.wikimedia.org/T137928 (10mmodell) [19:58:41] 10Phabricator, 10Release-Engineering-Team (Watching / External), 10Operations: Reimage both phab1001 and phab2001 to stretch - https://phabricator.wikimedia.org/T190568 (10mmodell) https://wikitech.wikimedia.org/wiki/Phabricator/Meeting_Notes/2019-01-23 [20:08:37] (03CR) 10Mainframe98: [C: 04-1] "> > it doesn't work […] as it checks against the current line only." [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [20:09:28] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:18:52] 10Release-Engineering-Team (Kanban), 10Education-Program-Dashboard, 10MediaWiki-extensions-EducationProgram, 10Epic, 10User-greg: Deprecate and remove the EducationProgram extension from Wikimedia servers after June 30, 2018 - https://phabricator.wikimedia.org/T125618 (10Zoranzoki21) Should this be close... [20:23:40] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:24:55] thcipriani mutante hi, can we also merge https://gerrit.wikimedia.org/r/c/operations/puppet/+/482379 ? (i asked volker who said he will not block the change). [20:25:28] It has 3 +1's (from the community) and i sent the email 2+ weeks ago so long enough to here feedback. [20:33:17] paladox: I missed your comment on that change, sounds like you're saying on the old ui we'll still see the old logo? [20:33:55] thcipriani nope, the old ui will just use png [20:33:58] instead of svg [20:34:06] where as the new ui will use svg (and not png) [20:34:15] as all browsers PolyGerrit supports, support svg [20:34:42] ah, ok. As long as the old ui has a logo, that was my primary concern. [20:35:02] lemme re-test [20:35:06] (03CR) 10Thiemo Kreuz (WMDE): "> It already has code to find out the last parenthesis of the if statement […]" [tools/codesniffer] - 10https://gerrit.wikimedia.org/r/486033 (https://phabricator.wikimedia.org/T203800) (owner: 10Thiemo Kreuz (WMDE)) [20:35:22] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:41:32] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:42:06] thcipriani ok [20:42:32] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:45:59] thcipriani https://gerrit-test.wmflabs.org/gerrit/q/status:open (works) [20:46:06] https://gerrit-test.wmflabs.org/gerrit/q/status:open?polygerrit=0 [20:50:31] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [20:59:34] 10Continuous-Integration-Config, 10Release-Engineering-Team (Kanban), 10Wikipedia-Android-App-Backlog: Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10hashar) Nice bug report, it got all the informations :) The Jenkins job clones the repository and applies... [21:08:49] 10Continuous-Integration-Config, 10Release-Engineering-Team (Kanban), 10Wikipedia-Android-App-Backlog: Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10hashar) The update of the license file on integration-slave-jessie-1002 at Jan 18 18:16 would correspond t... [21:12:27] thcipriani did you manage to test it? :) [21:22:38] 10Continuous-Integration-Config, 10Release-Engineering-Team (Kanban), 10Wikipedia-Android-App-Backlog: Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10hashar) Eventually I have found: https://discuss.circleci.com/t/android-platform-28-sdk-license-not-accept... [21:23:55] 10Continuous-Integration-Config, 10Release-Engineering-Team (Kanban), 10Wikipedia-Android-App-Backlog: Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10hashar) Beside the above diagnostic, I don't know what has to be fixed. The CI job a least is rather dumb:... [21:35:54] hey all, need some help with the new extension [21:36:58] do I need some special setup to get extension tests running on CI? [21:37:26] specifically, it's this one: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/WikibaseCirrusSearch/+/447954 (WikibaseCirrusSearch) [21:37:51] and from https://integration.wikimedia.org/ci/job/quibble-vendor-mysql-hhvm-docker/31875/consoleFull I'm not sure whether or not the extension tests actually run [21:38:10] also it depends on two other extensions, and I'm not sure whether I need to specify it somewhere [21:40:51] thcipriani, twentyafterfour, marxarelli ^^ ? [21:46:18] sure [21:46:22] * thcipriani reads [21:48:27] * thcipriani digging in integration/config [21:51:01] so the extension-quibble template brings in a whole bunch of tests for the repo: https://gerrit.wikimedia.org/r/plugins/gitiles/integration/config/+/master/zuul/layout.yaml#1202 [21:52:28] extension dependencies are defined in zuul/parameter_functions.py: https://gerrit.wikimedia.org/r/plugins/gitiles/integration/config/+/master/zuul/parameter_functions.py#139 [21:52:53] thcipriani: so that means I need to declare deps there? [21:54:00] right, that should get EXT_DEPENDENCIES filled in in the parameters listed: https://integration.wikimedia.org/ci/job/quibble-vendor-mysql-hhvm-docker/31875/parameters/ [21:54:26] weirdly enough, Wikibase declares dependency on WikibaseLexeme - shouldn't it be the reverse? [21:54:27] which will pull in extension dependencies before running your tests [21:54:40] ah, it declares both... makes sense I guess [21:55:40] thcipriani: ok, and then do I need to do anything to enable CI to acutally run unit test for this ext? [21:56:14] because I suspect right now it doesn't run them (also because it's green and I even didn't define the dependencies...) [21:56:37] so I wonder if I am missing some part there [21:56:52] that is suspicious, let me see what I can gleen from the quibble container [21:57:04] (I'm not familiar with quibble internals, honestly) [21:57:26] the test do pass locally, but I am not sure CI even runs them... I suspect not [21:59:53] (03PS1) 10Smalyshev: Add CI configurations for WikibaseCirrusSearch [integration/config] - 10https://gerrit.wikimedia.org/r/486170 (https://phabricator.wikimedia.org/T190022) [22:03:23] thcipriani: do you know what's the difference between extension-quibble-composer and extension-quibble ? [22:05:34] at a first pass quibble seems to run: php tests/phpunit/phpunit.php --debug-test --testsuite extensions --exclude-group Database [22:05:49] followed by a version of that command not excluding Database tests [22:07:01] thcipriani: but does it run the extension tests? the first run has too many tests, and the second too little... [22:07:07] so am not sure how to make sure [22:07:22] maybe I'll add some broken test to see [22:08:18] paladox: https://bugs.chromium.org/p/gerrit/issues/detail?id=10369 [22:08:43] thcipriani: also: https://gerrit.wikimedia.org/r/c/integration/config/+/486170 - does this look good? [22:09:36] * paladox looks [22:09:56] (03CR) 10Thcipriani: [C: 03+2] Add CI configurations for WikibaseCirrusSearch [integration/config] - 10https://gerrit.wikimedia.org/r/486170 (https://phabricator.wikimedia.org/T190022) (owner: 10Smalyshev) [22:09:59] lgtm [22:10:13] I wonder if doing just localStorage (if window is not defined) will work Krinkle [22:10:27] paladox: window is always defined. [22:10:34] window.localStorage is also always defined [22:10:41] hmm [22:10:58] accessing window.localStorage throws a permission exception. This is explicit, it is not an error. [22:11:22] window = { localStorage get() { if (notAllowed) { throw DOMException(''); } } [22:11:29] Works kind of like that. [22:11:31] https://github.com/GerritCodeReview/gerrit/search?utf8=%E2%9C%93&q=localStorage&type= [22:11:37] thcipriani: so I broke one unit test for this ext, and the CI still passes. So it doesn't run the tests for this extension [22:11:48] (03Merged) 10jenkins-bot: Add CI configurations for WikibaseCirrusSearch [integration/config] - 10https://gerrit.wikimedia.org/r/486170 (https://phabricator.wikimedia.org/T190022) (owner: 10Smalyshev) [22:12:01] so i take it your hitting https://github.com/GerritCodeReview/gerrit/blob/13975fa2aef6dae1bdf57fddbea10d02fca4c95e/polygerrit-ui/app/elements/gr-app.js#L144 Krinkle [22:12:04] that is..bizarre. [22:12:35] thcipriani: https://integration.wikimedia.org/ci/job/quibble-vendor-mysql-hhvm-docker/31896/console says success [22:12:58] SMalyshev: apologies, this is the first time I've looked at this space in a while and the world has moved forward quite a bit since the last time I looked. [22:13:25] thcipriani: no problem :) [22:13:41] paladox: Yes, the access to window.localStorage must be wrapped in try/catch. There are multiple ways to do that. For example, one could have a simpler class that wraps it (e.g. myStorage = { getItem: function() { try { return localStorage.getItem(..); } catch() {return null;}}}) [22:13:42] do you know who may know this better maybe? [22:13:56] ah thanks [22:13:59] so this one: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/WikibaseCirrusSearch/+/447954/14/tests/phpunit/Query/HasWbStatementFeatureTest.php#366 should have failed the test (and does locally) [22:14:07] but no failure on CI [22:14:15] i guess i could make a simple gr-local-storage behavour [22:14:23] Or one could define try { safeStorage = window.localStorage; } catch (e) { safeStorage = null; } and then everywhere go like if (safeStorage && safeStorage.getItem) [22:14:25] Lots of ways. [22:14:31] that try { ... } catch { return false; } [22:14:55] https://github.com/GerritCodeReview/gerrit/blob/d7116e12c587bf457c6e32b00b165da12a8289ce/polygerrit-ui/app/elements/shared/gr-storage/gr-storage.js [22:14:59] There is already something like that. [22:15:03] But, it is not yet used for dark mode. [22:15:09] And it also is missing try/catch for _value access. [22:15:14] oh [22:15:15] That also needs it, or will have the same error [22:15:15] * paladox looks [22:15:22] !log reloading zuul to deploy https://gerrit.wikimedia.org/r/#/c/integration/config/+/486170/ [22:15:23] Logged the message at https://wikitech.wikimedia.org/wiki/Release_Engineering/SAL [22:15:46] ^ SMalyshev job changes deployed [22:15:47] https://github.com/GerritCodeReview/gerrit/blob/master/polygerrit-ui/app/elements/shared/gr-storage/gr-storage.js this Krinkle ? [22:16:01] SMalyshev: hashar is the one who knows this best for sure [22:16:15] paladox: That is the same file, yes. [22:16:22] ok [22:16:23] thanks [22:17:14] thcipriani: ok if we don't figure it out I'll ask him then [22:23:38] ok, so let me see if I can figure out what command is being run here. so all quibble jobs run https://gerrit.wikimedia.org/r/plugins/gitiles/integration/config/+/master/jjb/mediawiki.yaml#225 in the case of vendor-mysql-hhvm-docker it's running with https://gerrit.wikimedia.org/r/plugins/gitiles/integration/config/+/master/jjb/mediawiki.yaml#588 image [22:25:00] 10Continuous-Integration-Config, 10Release-Engineering-Team (Kanban), 10Wikipedia-Android-App-Backlog: Gate pipeline build failed on apps/android/wikipedia - https://phabricator.wikimedia.org/T214497 (10cooltey) Thanks @hashar! I found a related PR talking about the same issue: https://github.com/expo/expo/p... [22:25:27] Krinkle https://gerrit-review.googlesource.com/c/gerrit/+/211472/ [22:25:30] which runs quibble, I guess https://gerrit.wikimedia.org/r/plugins/gitiles/integration/config/+/master/dockerfiles/quibble-jessie/Dockerfile.template#88 [22:26:36] yeah that seems to run generic extensions test, which for some reason does not include my extension... Maybe because it's not in WMF install yet? [22:27:06] but there should be some provision for testing non-WMF-prod extensions... so maybe I am using wrong magic there... [22:28:44] quibble --packages-source vendor --db mysql --db-dir /workspace/db --skip npm-test more specifically [22:29:05] well, I assume somewhere in quibble it's using the job parameters to download your patch [22:29:24] what --packages-source vendor means? [22:30:04] whole quibble is very new for me... [22:30:06] uses https://gerrit.wikimedia.org/r/plugins/gitiles/mediawiki/vendor/ instead of composer [22:30:24] ahh... so this ext is not there yet. Maybe that's the reason? [22:30:54] it's not on packagist either though... so not sure if that's something related to it [22:32:50] paladox: nice [22:33:07] now im trying to figure out how i switch my cookies off to test [22:33:30] ah [22:33:31] found it [22:33:42] hrm, well, it looks like it should clone mw/core, mw/vendor, as well as your project: https://gerrit.wikimedia.org/r/plugins/gitiles/integration/quibble/+/master/quibble/cmd.py#203 [22:34:02] ZUUL_PROJECT is set in https://integration.wikimedia.org/ci/job/quibble-vendor-mysql-hhvm-docker/31896/parameters/ to mediawiki/extensions/WikibaseCirrusSearch [22:39:21] which seems to work based on console output [22:39:29] > INFO:zuul.CloneMapper: mediawiki/extensions/WikibaseCirrusSearch -> extensions/WikibaseCirrusSearch [22:40:14] then it runs composer --ansi test in your extension directory [22:40:56] Krinkle i have another place to update [22:41:08] but cannot use the gr-storage module as it's outside of the dom [22:41:44] paladox: maybe make it a normal JS interface and use it both inside gr-storage and inside the dark mode init code. [22:42:16] it's only needs updating in one place [22:50:11] thcipriani: but composer test does not run phpunit [22:50:30] right [22:50:49] so I wonder what happens with the phpunit part :) [23:01:52] SMalyshev: phew, ok, I think this might be roughly it: https://gist.github.com/thcipriani/ee6e389a7d169c371a3235b2e2724e50 [23:01:59] translated from python back to shell [23:02:56] ok so I imagine this means my ext is not part of testsuire extensions for some reason? [23:03:34] looks that way [23:04:05] twentyafterfour: thcipriani: follow-up 1 https://phabricator.wikimedia.org/T195623#4903960 follow-up 2 https://gerrit.wikimedia.org/r/#/c/operations/puppet/+/486181/ [23:05:28] SMalyshev: not quite sure how looking at https://doc.wikimedia.org/mediawiki-core/master/php/ExtensionsTestSuite_8php_source.html but that seems like a more concrete problem than "what is this test running?" at least [23:05:57] mutante: awesome [23:05:59] :) [23:07:01] hmm maybe something missing in extensions.json... I'll try to read the code and figure out what it wants [23:08:34] seems to assemble *Test.php from /tests/phpunit - so it should work... [23:08:57] unless the registry does not pick the ext at all [23:09:57] 10Gerrit, 10Release-Engineering-Team (Next): Upgrade to Gerrit 2.15.9 - https://phabricator.wikimedia.org/T214359 (10greg) Looking like we'll want to do this after All Hands. [23:13:18] 10Release-Engineering-Team (Watching / External), 10Education-Program-Dashboard, 10MediaWiki-extensions-EducationProgram, 10Epic: Deprecate and remove the EducationProgram extension from Wikimedia servers after June 30, 2018 - https://phabricator.wikimedia.org/T125618 (10greg) [23:21:35] 10Release-Engineering-Team (Kanban), 10Patch-For-Review, 10Release, 10Train Deployments: 1.33.0-wmf.14 deployment blockers - https://phabricator.wikimedia.org/T206668 (10mmodell) [23:28:53] thcipriani: so I am trying to figure out what controls what extensions are loaded in CI and I don't see anything obvious there. Do you know what controls it? Because it looks like my ext is just not getting loaded at all [23:30:12] Krinkle got it working now [23:49:45] SMalyshev: hrm, looking at LocalSettings.php from the test you linked (in build artifacts) I don't see wmfLoadExtension anywhere. I am unclear how extensions are registered for that job.