[00:14:51] <wikibugs>	 (03PS1) 10Legoktm: Don't run special codesniffer handling for other libraries [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396649
[00:23:20] <wikibugs>	 (03CR) 10Legoktm: [C: 032] Don't run special codesniffer handling for other libraries [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396649 (owner: 10Legoktm)
[00:23:34] <wikibugs>	 (03Merged) 10jenkins-bot: Don't run special codesniffer handling for other libraries [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396649 (owner: 10Legoktm)
[00:33:03] <wikibugs>	 (03PS1) 10Legoktm: Allow limiting how many repositories run at a time [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396656
[00:33:15] <wikibugs>	 (03CR) 10Legoktm: [C: 032] Allow limiting how many repositories run at a time [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396656 (owner: 10Legoktm)
[00:33:30] <wikibugs>	 (03Merged) 10jenkins-bot: Allow limiting how many repositories run at a time [labs/libraryupgrader] - 10https://gerrit.wikimedia.org/r/396656 (owner: 10Legoktm)
[04:08:49] <zhuyifei1999_>	 bd808: what happened to k8s?
[04:09:15] <zhuyifei1999_>	 (regarding your comment about slurm)
[04:11:16] <bd808>	 zhuyifei1999_: nothing, but I do not think that kubernetes will ever be a one to one replacement for grid engine
[04:11:37] <zhuyifei1999_>	 well, it won't
[04:12:08] <bd808>	 I would like to see *all* webservices move to k8s, but there will be ad hoc jobs that are never a good fit for k8s
[04:12:38] <bd808>	 slurm seems to be a promising replacement for grid engine
[04:12:49] <zhuyifei1999_>	 but personally I think maintaining only one service (k8s) will be better than having two (slurm and k8s)
[04:13:56] * zhuyifei1999_ looks into slurm
[04:14:00] <bd808>	 kubernetes is awesome, but it is a container orchestration system designed to support microservice architectures
[04:14:11] <zhuyifei1999_>	 yeah right
[04:14:21] <bd808>	 that is maybe half of our Toolforge usage
[04:14:47] <bd808>	 one off jobs and cron jobs are the other half
[04:15:31] <bd808>	 slurm is a distributed job scheduler much like grid engine, but a bit more modern under the covers
[04:16:19] <bd808>	 it knows about things like cgroups and has quite a bit of flexibility in adding plugins to tune how the grid works
[04:17:08] <bd808>	 we have talked to quite a few university computing centers who have replaced their grid engine deploys with slurm over the last couple of years
[04:18:56] <zhuyifei1999_>	 cool
[04:19:01] * zhuyifei1999_ looks forward
[04:19:28] <bd808>	 I don't know when we might start on that project. :) there are so many things that need fixing
[04:20:26] <bd808>	 I think we will probably upgrade Grid Engine first so we can get rid of Trusty
[04:20:35] <zhuyifei1999_>	 before trusty goes away? we don't want to build a new jessie cluster for grid do we
[04:20:50] <zhuyifei1999_>	 hmm
[04:22:22] <bd808>	 April 2019 sounds like it is far away, but it really isn't that long
[04:23:05] <bd808>	 we have a couple of other big projects that need to be done by then too
[04:23:13] <zhuyifei1999_>	 k
[04:24:10] <bd808>	 until we finish the OpenStack Neutron migration I'm not going to make many promises about future timelines and priorities though
[04:24:44] * bd808 crosses fingers and hopes for <6 more months on that giant project
[04:32:34] <zhuyifei1999_>	 bd808: oh, when you have time, could you have a look at https://gerrit.wikimedia.org/r/394923
[04:33:17] <bd808>	 zhuyifei1999_: yes! Its on my list of things for the coming week
[04:33:24] <zhuyifei1999_>	 ok thanks
[04:34:06] <bd808>	 a week of "working vaction" has my next week looking pretty horrible, but I think this should be relatively easy to test
[04:35:42] <zhuyifei1999_>	 k
[04:38:58] <bd808>	 where is the magic page on mediawiki.org that lets you get added as a reviewer automatically for a gerrit repo or path in a repo?
[04:39:28] <bd808>	 oh... it's actually in gerrit itself
[04:41:32] <bd808>	 nope -- https://www.mediawiki.org/wiki/Git/Reviewers
[05:49:30] <zhuyifei1999_>	 !log quarry quarry-main-01: `ALTER IGNORE TABLE star ADD UNIQUE INDEX star_user_query_index (user_id, query_id);` Records: 728  Duplicates: 17  Warnings: 0  T165169
[05:49:34] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Quarry/SAL
[05:49:34] <stashbot>	 T165169: Some querries cannot be 'unstarred' - https://phabricator.wikimedia.org/T165169
[05:52:43] <zhuyifei1999_>	 !log quarry deployed e835a46 to quarry-main-01 and restarted uwsgi T165169
[05:52:46] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Quarry/SAL
[16:21:24] <d3r1ck>	 bd808: Hello :)
[16:22:03] <d3r1ck>	 Did some updates to the doc here: https://wikitech.wikimedia.org/wiki/Help:Access#Connection_closed_by_remote_host
[16:37:42] <d3r1ck>	 Hope they make some sense? :)
[16:42:16] <GoranSM>	 Hey, do we have a replica of the log database that can be accessed from a Labs instance somewhere?
[17:46:08] <zhuyifei1999_>	 GoranSM: yes, the wiki replicas should be accessible from all labs instances
[17:46:58] <Zppix>	 Cloud*
[18:22:28] <eranroz>	 hi, I need sqlclient for a kubernetes service  (based on python3 image), how to install there mysql client? alternatively, should I use different image?
[18:26:25] <zhuyifei1999_>	 eranroz: on toolforge k8s?
[18:26:48] <zhuyifei1999_>	 Zppix: yeah
[18:41:19] <eranroz>	 yes
[20:29:00] <wm-bot>	 I will let you know when I see eranroz and I will deliver that message to them
[20:29:00] <zhuyifei1999_>	 @notify eranroz it's only installed on python2 containers per https://phabricator.wikimedia.org/T140112. for python3 you can use alternatives like pymysql