I would like to know if there is a way to specify the nodes Kibana connects to to perform the requests.
In particular I would like to set something similar to
preference=_local to all the queries executed by Kibana.
The idea is to use some nodes with enough resources to perform analysis (heavy aggregations) leaving unloaded other nodes used for data collection and light searches.
It's the exact opposite of Configure Kibana for multiple ES servers/nodes.