[09:09:21] <jayme>	 ottomata: IIRC we discussed that before. There is no normal deploy process for staging-codfw (as that's an SRE playground usually). But if it helps I can undeploy eventgates there
[13:46:45] <elukey>	 hello folks, after a chat with Janis I tried to improve https://wikitech.wikimedia.org/wiki/Kubernetes/Resource_requests_and_limits with some details and a new use case
[13:47:05] <elukey>	 claime: o/ let me know if you see/read anything weird in there
[13:47:16] <elukey>	 I tried to add more details based on previous discussions
[13:47:35] <elukey>	 specific diff: https://wikitech.wikimedia.org/w/index.php?title=Kubernetes%2FResource_requests_and_limits&diff=2127134&oldid=2124815
[14:20:04] <claime>	 elukey: thanks, will take a look
[14:20:42] <inflatador>	 Many thanks elukey , I'm always confused by this
[14:25:40] <elukey>	 I hope I get it right with the 5ms slices etc..
[14:26:05] <elukey>	 and there is always room for improvements, especially as we get experience with workloads etc..
[14:47:05] <ottomata>	 jayme:  i remember discussing it before, and that makes sense, but we shouldn't leave old code running there.  The errors are happening because we deprecated a stream config  API feature that old eventgate code was using.  we fixed it in all the normal places.
[14:47:15] <ottomata>	 jayme:  it would help to undeploy them there, thank you
[14:49:13] <claime>	 elukey: The slice calculation and stuff look all right to me, and you added a whole bunch of details! Thanks!
[14:49:45] <elukey>	 claime: <3
[14:55:12] <ottomata>	 elukey:  <3 wow very nice.  i was really missing the translation between millicores and ms cpu time.  
[14:55:14] <ottomata>	 ty
[15:20:24] <elukey>	 <3
[16:18:54] <jayme>	 ottomata: I've removed all eventgates from staging-codfw
[16:24:19] <ottomata>	 thank you!