[10:13:01] <gehel>	 errand, back in a bit
[15:32:41] <ebernhardson>	 \o
[15:57:51] <ebernhardson>	 started up mjolnir, not expecting craziness but will be monitoring how it loads the clusters today
[17:51:42] * ebernhardson is surprised that pyarrow FileSystem has a 'delete_root_dir_contents' method
[18:14:34] <ebernhardson>	 unrelatedly, increasing the search poolcounter by 10% on 11/17 seems to have done nothing for rejections :( if anything there are more pool counter rejections since then. If we actually want to deal with that going to have to dive into some analysis 
[18:15:22] <ebernhardson>	 my hunch is automated queries...but no proof :P
[18:18:30] <ebernhardson>	 but the query load from mjolnir so far seems plenty managable
[18:20:57] <ebernhardson>	 codfw can see a small bump in search threadpools, in eqiad it's lost in the natural variation
[19:03:20] <ryankemper>	 ebernhardson: gehel: need to cancel pairing today, appt’s getting quite delayed
[19:03:27] <ebernhardson>	 kk
[19:08:45] <gehel>	 ryankemper: ack