I'm tuning my ingestion pipeline to maximize throughput. I've read through the recommendations in the reference guide about how to tune for index speed (https://www.elastic.co/guide/en/elasticsearch/reference/master/tune-for-indexing-speed.html)
With all the different settings to experiment with, I find myself wondering what is going on behind the scenes during a bulk index call. What stages are there, how is work divided up, and what is parallelized?