Fast load utility is used to load the data in to empty table from a file with highspeed .

 Tip : In general fast load is used for loading to volatile table and then to final target table to enhance the overall performance .

 Two Phases :
Acquisition Phase :
 PE sends Blocks of data to Each AMP in round robin.
 AMPS hash and re-distributes to appropriate AMPs.
 Records that are  written to AMP spool file .the records are not sorted.
Application Phase :
This Phase starts once END LOADING is encountered
 Every AMP sort spool file and creates data blocks.
 Writes data blocks to disk.
 Fallback rows generated if required.
Limitations:
Target Table should be empty
No Triggers are allowed
No Referential Integrity is allowed ( Foreign Key Relationship)
Duplicate Rows are not supported

Secondary Index is not allowed