In the existing hive target, we have an option “overwrite partitions” which will perform the upsert operation on the target table as per the source data.

Dataflow à Target instance à properties à Overwrite Partitions.

When the overwrite partitions is selected it will insert and overwrite the data present in already existing hive table level partitions. It will not perform entire table level insert and overwrite. Which is a performance improvement during the execution in this scenario.