I noticed when investigating slow performance of the map on a dev site using the map/reduce system, that the performance in my IDE on my laptop was around 6.5 seconds while the map was about 9.5 seconds.
We should reproduce and document these experiments to make sure the results are valid/correct. This could also involve enabling the Java Flight Recorder temporarily in our dev environments.
Here is the request in our largest restricted study:
Overview
I noticed when investigating slow performance of the map on a dev site using the map/reduce system, that the performance in my IDE on my laptop was around 6.5 seconds while the map was about 9.5 seconds.
We should reproduce and document these experiments to make sure the results are valid/correct. This could also involve enabling the Java Flight Recorder temporarily in our dev environments.
Here is the request in our largest restricted study: