[08:21:13] <RhinosF1>	 lmata: https://github.com/Ladsgroup/Phabricator-maintenance-bot/pull/36
[08:21:21] <RhinosF1>	 cc Amir1 
[09:13:57] <Amir1>	 RhinosF1: {{merged}}
[09:15:49] <RhinosF1>	 It's nice when I disappear and magic happens
[11:06:19] <Amir1>	 no magic involved, I just like hitting shiny buttons
[11:25:47] <RhinosF1>	 It feels like magic
[11:27:27] * RhinosF1 should be on lunch
[11:27:46] <RhinosF1>	 It was quiet when people were asleep in real life
[13:25:41] <majavah>	 hi all. the elk7 setup in deployment-prep is complaining about "Field [timestamp] of type [text] does not support custom formats", any hints where to look at?
[13:26:19] <majavah>	 and second question: should I just have one "logstash*" index pattern, or use separate patterns for mediawiki, deploy, syslog, so on?
[14:27:04] <lmata>	 Thanks RhinosF1 I had that in my todo much appreciated
[14:28:19] <RhinosF1>	 lmata: Happily!
[15:01:07] <cwhite>	 majavah: the custom formats error message sounds like there is a producer sending timestamps in an unrecognizable format
[15:03:16] <cwhite>	 majavah: the separate index patterns for those types you identified are mitigations to deal with field type conflicts.  ideally, we wouldn't need separate patterns but if you're seeing a situation where important logs are getting dropped due to type conflicts, the separate patterns is a way to cope with that.