Hi Hubert, will this work with all DML operations and on all Spark versions? And can i use NUM_AFFECTED_ROWS for all DML operations?