[08:50:26] good morning! coming back to my request from yesterday... is anyone around who would be able to give a +1 on gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1354040 , be around while I deploy it, and then deploy https://gerrit.wikimedia.org/r/c/operations/puppet/+/1356371 ? [09:01:03] effie, jayme, are you around? [09:03:53] some of us are out today (incl me), that's why I suggested Wed EU morning :) [09:04:47] probably janis will be around soonish [09:06:18] Oh, I missed the "Wed" part :) [09:07:06] Ok, if nobody volunteers today, I'll bug you again tomorrow... [10:26:32] sorry, I'm not fully catched up from the offsite week. If it can wait until tomorrow that would be great [10:27:52] from clicking on it it seems there is a CI change bundled with this. I would suggest to split that up into a parent commit to seperate concerns [10:54:12] jayme: the test change tests the new routes. I could factor out a baseline, for testing the old routes. But deploying it separately doesn't make sense imho. Note that these are deployment tests, not CI tests. They run against staging after deployment. [10:54:34] ...manually [10:54:58] I was talking about this https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1354040/4/.rake_modules/tester/asset.rb [10:55:40] ah, right. yes, I can split that out. I did that because rake failed for me when validating locally. [10:56:07] ah, I see, newer rake version probably. [10:59:05] hello, in puppet I need to change the shebang for the mediawiki-firejail-ffmpeg and mediawiki-firewall-ghostscript scripts from "python" to "python3" to please CI/Rake , else it lint them as python2 [10:59:42] I am not sure whether those scripts are still used, presumably the functionality got moved to Shellbox. Though maybe that is still used on deployment-prep [10:59:57] the change is tentatively straightforward: s/python/python3/ https://gerrit.wikimedia.org/r/c/operations/puppet/+/1364525 ;) [11:00:28] and if it still runs somewhere, it is almost certainly already running as python3 [11:12:31] jayme: updated [14:35:59] Hi, it's me again. I made some changes based on Hugh's feedback but we need someone from serviceops to sign off the resource changes: https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1367963 [14:39:28] taking a look [14:41:50] Thanks! [14:42:19] The backlog is at 8K videos and I haven't even started the full transcode of all videos to DASH [14:46:00] i'm a little unclear about what's meant to consume the "job_overrides" stanza [14:46:25] (in helmfile.d/services/mw-videoscaler/values.yaml) [14:47:14] https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1367963/2/helmfile.d/services/mw-videoscaler/values.yaml to basically allow the normal workers pick stuff from prio queue too [14:47:40] (and increase the number of workers) [14:49:14] sorry, what i mean is, looking through charts/mediawiki/templates/job.yaml.tpl, i don't see a "job_overrides" changing the configuration of the job that would be created here [14:51:48] ah this must be me misunderstanding how this is templated [14:52:01] looking through the diff, it looks like it does the intended thing [14:53:47] https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1367963/2/charts/mediawiki/templates/mercurius/configmap.yaml.tpl [14:54:51] YAMLs like this reduces my will to live [14:54:51] oh, yes, that bit :D [14:55:05] yes this is Big Pain [14:55:08] looks good to me, though [14:55:38] Thanks <3 [14:55:44] np!