extended_stats aggregations provides a bunch of metrics. Is it possible to define/filter which of these metrics I need returned?
I have a million documents and I need standard deviation metric only. I want to exclude any performance hit on computing other extended metrics that I don't need.
A lot of statistics, like the standard deviation are derived from the average, so you can't actually calculate one without the other. Rather than trying to avoid any performance hit, it might be worth defining what is an acceptable response time in your application and make sure your searches take less than that.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.