i'm looking for a way for elastic to generate a spark job which will perform aggregations and write them back to elasticsearch and for them to be available for kibana to display.
we're using kibana as a BI platform for our data, but it's getting too massive.
we can increase the compute (by a lot) but the costs will be very high.
so i'm looking for a way to keep the raw data on S3 and for it to be aggregated with a job invoked by elastisearch on demand. then, the results will be written to elasticsearch and displayed in kibana.
is that possible?