[00:20:55] <wm-bot>	 I will let you know when I see framawiki and I will deliver that message to them
[00:20:55] <zhuyifei1999_>	 @notify framawiki np ;)
[01:43:59] <Dispenser>	 I have to wonder how many Wikipedians only get informed of server issues/replag because of the notice on my site
[01:46:37] <Vermont>	 Dispenser: what’s your site
[01:46:54] <Dispenser>	 http://69.142.160.183/~dispenser/view/Main_Page
[01:47:26] <Dispenser>	 It got rdcheck (linked from WhatLinksHere) and disambiguation tools
[01:48:24] <Vermont>	 Added to my bookmarks. Nice site.
[01:49:16] <Dispenser>	 Thanks
[09:31:59] <chrismytton>	 !help I'm trying to get a Ruby application running on Kubernetes, but I'm seeing an error about not being able to find `libmysqlclient.so`, I think because the "ruby" docker container doesn't have the libmysqlclient-dev package installed
[09:32:00] <wm-bot>	 chrismytton: If you don't get a response in 15-30 minutes, please create a phabricator task -- https://phabricator.wikimedia.org/maniphest/task/edit/form/1/?projects=wmcs-team
[09:32:28] <chrismytton>	 I notice that the "python2" docker container does have this package - https://phabricator.wikimedia.org/diffusion/ODIT/browse/master/python2/base/Dockerfile.template;bac775da7cbeb7c275ae0a6251e9401524b6a2c0$18
[09:33:05] <chrismytton>	 so does that package need to be added to https://phabricator.wikimedia.org/diffusion/ODIT/browse/master/ruby/base/Dockerfile.template, or is there another workaround?
[09:34:41] <chrismytton>	 I'm trying to start the service with `webservice --backend=kubernetes ruby2 start /data/project/verification-pages/verification-pages/script/toolforge-start`, you can see that script here: https://github.com/mysociety/verification-pages/blob/d03f565b1f0bf0f03e07a125746c61e49a15b0d4/script/toolforge-start
[09:36:02] <chrismytton>	 here's a gist of the full error: https://gist.github.com/chrismytton/fd1e254a393ebd36be0b5d9245138aab
[10:50:31] <arturo>	 chrismytton: yeah, likely a missing library. I would say the way to go is toopen a phabricator task asking for it
[11:34:31] <chrismytton>	 arturo: OK I'll do that. Many thanks!
[14:53:06] <wikibugs>	 (03PS1) 10Ottomata: Add certificates for kafka_test_broker and kafka_main-deployment-prep_broker [labs/private] - 10https://gerrit.wikimedia.org/r/427676 (https://phabricator.wikimedia.org/T167039)
[14:56:50] <wikibugs>	 (03CR) 10Ottomata: [V: 032 C: 032] Add certificates for kafka_test_broker and kafka_main-deployment-prep_broker [labs/private] - 10https://gerrit.wikimedia.org/r/427676 (https://phabricator.wikimedia.org/T167039) (owner: 10Ottomata)
[15:46:33] <andrewbogott>	 Krinkle: do you have a moment?  I want to apply https://gerrit.wikimedia.org/r/#/c/427690/  but I see a hack that you added in noc/db.php that checks for the old 'labtestweb' cluster which will no longer exist after my patch.
[15:46:44] <andrewbogott>	 I'd like to keep that hack in place but switch on the db name rather than the cluster… is that easy?
[15:48:49] <Reedy>	 if ( $i === 0 && strpos( $clusterName, 'labtestweb' ) !== 0 ) {
[15:51:00] <andrewbogott>	 Reedy: ok then :)
[15:51:22] <Reedy>	 that's what's there currently
[15:51:30] <Reedy>	 so $dbname === or in_array() or similar
[15:51:32] <andrewbogott>	 ah, yes, I misread
[15:52:48] <andrewbogott>	 Reedy: that function doesn't know the dbname does it?
[15:52:54] <Reedy>	 The loop does
[15:52:57] <Reedy>	 Or, should
[15:53:06] <Reedy>	 			foreach ( $this->getDBs( $clusterName ) as $i => $db ) {
[15:53:07] <Reedy>	 				print "$db";
[15:53:07] <andrewbogott>	 oh, I see
[15:53:13] <andrewbogott>	 ok, trying...
[15:53:47] <andrewbogott>	 I totally don't understand what that $i === 0 part is doing
[15:54:52] <Reedy>	 array indexes
[15:55:03] <Reedy>	 so, literally the first item in the array returned by getDBs
[15:56:35] <andrewbogott>	 if ( $i === 0 && $db !== "labtestwiki" ) {
[15:56:43] <andrewbogott>	 That's all I need there?
[15:57:03] * andrewbogott unsure about !== vs !=
[15:59:44] <andrewbogott>	 Reedy: https://gerrit.wikimedia.org/r/#/c/427690/
[16:51:57] <Krinkle>	 andrewbogott: != will allow typecasting, !== will not
[16:52:22] <Krinkle>	 e.g. 0 == '0', and even '0' == false (crazy), but not with ===
[17:12:26] <andrewbogott>	 Krinkle: I sort of know that but still don't really understand what it will do with strings.  Anyway, thanks for the review!
[17:12:39] * andrewbogott looks at deployment calendar
[17:13:09] <Krinkle>	 andrewbogott: For plain strings (not objects that have a toString), == and === would do the same. But as general principle, just always use === and cut out the uncertainty all together :)
[17:13:23] <Krinkle>	 Because unless the code is 'string' === 'string', you can never be entirely sure whetever $x is or is not a string.
[17:13:30] <andrewbogott>	 Krinkle: That's fair :)
[19:36:23] <andrewbogott>	 !log editor-engagement deleting project as per T190952
[19:36:24] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Editor-engagement/SAL
[19:36:24] <stashbot>	 T190952: Delete VPS project 'editor-engagement' on or after 2018-04-15 - https://phabricator.wikimedia.org/T190952
[19:39:59] <chasemp>	 andrewbogott: hey man I'm sorry for missing the last round of project creations during my clinic, the week got away from me and then I wasn't around to catch up mon/tue so thank you
[19:40:19] <andrewbogott>	 no problem — I think we didn't approve anything in any case
[19:47:23] <wikibugs>	 (03Abandoned) 10Paladox: Fix gerrit not so it dosen't play jenkins-bot comments [labs/tools/wikibugs2] - 10https://gerrit.wikimedia.org/r/424965 (owner: 10Paladox)
[20:40:09] <andrewbogott>	 ragesoss: what can you tell me about outreachdashboard.globaleducation.eqiad.wmflabs?  It sure it chewing up CPU cycles
[20:40:31] <andrewbogott>	 and has been for days and days without interruption as best I can tell
[21:43:15] <zhuyifei1999_>	 !log toolsbeta Start creating instances for webservice setup T190893
[21:43:16] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL
[21:43:16] <stashbot>	 T190893: Setup the webservice-related instances in toolsbeta - https://phabricator.wikimedia.org/T190893
[21:45:07] <zhuyifei1999_>	 andrewbogott: I can't seem to create trusty instances. Will there be an issue if they are created in jessie?
[21:45:18] <zhuyifei1999_>	 oge probably won't work
[21:48:59] <andrewbogott>	 > Will there be an issue if they are created in jessie
[21:49:09] <andrewbogott>	 zhuyifei1999_: what does 'they' refer to there?
[21:50:27] <zhuyifei1999_>	 the toolforge webservice replicated setup for T190893. Some 'must needed' instances like bastion and OGE are running on trusty instances
[21:50:28] <stashbot>	 T190893: Setup the webservice-related instances in toolsbeta - https://phabricator.wikimedia.org/T190893
[21:50:29] <andrewbogott>	 oh, grid nodes
[21:50:48] <andrewbogott>	 zhuyifei1999_: I enabled the Trusty base for tools-beta; do you see it now?
[21:51:06] <zhuyifei1999_>	 ok thanks, will check. doing the jessie ones right now
[21:51:38] <zhuyifei1999_>	 yeah I see, thanks again
[21:51:51] <andrewbogott>	 great
[21:57:32] <zhuyifei1999_>	 !log toolsbeta Add proxy toolsbeta.wmflabs.org => toolsbeta-proxy-01.toolsbeta.eqiad.wmflabs
[21:57:34] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL
[21:58:21] <zhuyifei1999_>	 what is 'Error: RecordSet belongs in a child zone: toolsbeta.wmflabs.org'
[21:59:15] <zhuyifei1999_>	 !log toolsbeta got 'Error: RecordSet belongs in a child zone: toolsbeta.wmflabs.org', using tools-beta.wmflabs.org instead
[21:59:15] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL
[22:10:47] <zhuyifei1999_>	 !log toolsbeta the trusty instances ask me for my password. the jessie instances don't like my ssh key. :(
[22:10:48] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL
[23:32:01] <chicocvenancio>	 Not strictly WMCS related, but we have a negotiation with a GLAM partner for arround 8TB of images, is server side upload as described in https://commons.wikimedia.org/wiki/Help:Server-side_upload#What_to_do_if_files_represent_hundred_of_GB_to_several_TB%3F Still the best way to go? 
[23:50:53] <zhuyifei1999_>	 chicocvenancio: yes
[23:52:09] <chicocvenancio>	 Thanks zhuyifei1999_ 
[23:52:17] <zhuyifei1999_>	 unless they can find some way to host that 8TB of images... with some easy way for the 'deployers' to fetch them
[23:52:38] <zhuyifei1999_>	 (which sounds unlikely and might take more time than sending a hard drive)
[23:53:54] <zhuyifei1999_>	 chicocvenancio: oh could you check the instances I just created two hours ago on toolsbeta
[23:54:08] <zhuyifei1999_>	 the majority of them looks broken (don't allow me to login)