A data engineer runs the following query to load the data in the downstream table. spark . readStream . format ( 'delta' ) . table ( 'upstream' ) . writeStream . option ( 'checkpointLocation' , '/tmp/upDown' ) . format ( 'delta' ) . trigger ( once = True ) . table ( 'downstream' ) After a few days, the upstream table was dropped and re-created as one of the columns needed to be removed to be compliant with the data protection rules. As new data arrived in the upstream table, the query was re-run but it failed. Which of the following changes should be done to the query to run it successfully keeping the historical data intact in the downstream table?