Hello Elastic team,
I recently got challenges on a complex aggregation search that when this aggregation is run sequentially by a python script, after some time goes on, it starts receiving circuit breaker errors on search phase. At this flow it's acceptable that many sequential complex aggregations can lead to jvm heap high load and cause circuit breaker comes in! But for some of my entry sources even executing a single aggregation request gets no answer and again raises the circuit breaker error. But when I execute that aggregation with 'async_search' API, it's no pain for ES to compute the calculations and the response is ready after some time. My question is that does async_search in the background performs more optimized than the regular search API in case of resource consumption?