[06:49:04] <tgr>	 is global user merge still a thing on WMF wikis?
[06:53:51] <legoktm>	 no
[06:54:20] <legoktm>	 but the code is still somewhat deployed because I haven't cleaned it up yet :(
[22:43:11] <legoktm>	 Reedy: hi
[22:43:18] <legoktm>	 What happened?
[22:43:28] <Reedy>	 legoktm: https://phabricator.wikimedia.org/T148609#3184036
[22:43:36] <Reedy>	 Linter was creating a *huge* amount more db queries
[22:43:54] <legoktm>	 I see
[22:44:06] <Reedy>	 Was especially causing issues on commonswiki
[22:44:13] <Reedy>	 So it's been rolled back on all large wikis
[22:44:29] <legoktm>	 That's fine
[22:44:48] <legoktm>	 I think parsoid needs to be restarted for it to notice that linter is gone
[22:45:00] <Reedy>	 Nothing furhter to stop that warning spam?
[22:45:07] <Reedy>	 I guess they load the mw config?
[22:45:24] <legoktm>	 It loads meta=siteinfo at startup
[22:45:36] <legoktm>	 So a rolling restart of the cluster should fix it