We want to copy a large Spark dataframe into Oracle, but I am finding the tuning options a bit limited.