[11:36:38] I have a +1 from marostegui on re-enabling multi-DC https://gerrit.wikimedia.org/r/c/operations/puppet/+/824039 [11:36:47] I cherry-picked and deployed the fixes [11:36:56] so that can go out tomorrow my time [12:50:53] Amir1: ApiQuery/ApiQueryBase ::selectNamedDB/::getNamedDB - does it bring joy? [12:51:14] Krinkle: nope, only pain and suffering [12:51:28] removal incoming [12:51:30] deprecation* [12:51:36] Thanks! [13:12:23] AaronSchulz: would you consider this solved? ref T269325 [13:12:23] T269325: Implement a reasonable strategy for handling Excimer-style timeouts in MediaWiki - https://phabricator.wikimedia.org/T269325 [13:26:29] Krinkle: T277927 has SRE tag but no owner nor priority, it is my duty as clinic duty this week to be annoying and ask for a way to remedy this, the patch looks good to me, should we merge it? [13:26:30] T277927: Add monitoring for performance.wikimedia.org - https://phabricator.wikimedia.org/T277927 [13:31:07] Amir1: I don't know the 'blackbox' class or how it works I'm guessing it tells Icinga to run a script (from where?) at some interval (what interval?) to check GET $server_name/ over some schema (HTTPS?), and then report severity 'critical' which I guess is lower than 'page', i.e. IRC only? [13:31:48] if there's an easy way for this to go to the perf contact group (#wikimedia-perf- bots + perf-team-alerts) that would be great. [13:31:55] but LGTM either way as starting poitn [13:32:07] hmm, I will see what I can do about that [13:55:20] yeah basically what Daniel suggested is the way to go now, i.e. via blackbox checks through prometheus/alertmanager (no icinga) [18:00:25] Krinkle, lgaulia: does the google meet link work for you? [18:58:27] Krinkle: I think https://phabricator.wikimedia.org/T269325 would need some auditing/work before closing [20:48:10] Krinkle: I amended https://gerrit.wikimedia.org/r/c/mediawiki/core/+/790467/