When you run a mapping on the Blaze engine to write data to a flat file target, the Data Integration Service creates multiple target files with the following naming convention:

<FileName>-P1, <FileName>-P2,....<FileName>-P100....<FileName>-PN

The naming convention helps you to delete multiple split files generated from the previous mapping runs and to avoid the deletion of target files generated from another mapping with similar file names.