We have over 25 million GPS data that we need to index on a daily basis usig Spark. We use the geoshape datatype with the LINESTRING feature. We are finding that indexing performance is extremely slow (on the order of minutes for a small sample) using the bulk api, esp in comparison to postgres. Any tips on how to improve indexing performance would be appreciated. We've disabled replicas as well as index refresh until ingestion is finished.
Thanks,,
Trang