Hello - Could someone please tell me what the load utility does in cases where rows have been modified in a table after an export has been done? You issue the load command with insert into, how are those modified rows handled ?

Thanks

I think there may be some misunderstanding on your part.

If you do an export, the data is stored in a flat file (you specify the file name in the export command). Then when you do a load (or import) at some later time interval, the flat file with the name you specify is used as input and the rows are added to the table you specify.

Right, I'm aware of this. My question is does the data in the file overwrite the changes in the table if you load the file into the same table at a later time using db2 load?

If you load the data back into the same table?

• If there is no unique index (or PK) on the table, it will add the row, leaving the old row in place.
• If there is a unique index or PK on the table, then the new row will be rejected during the load process (it will show up in the Deleted count in the output) becaue it will violate a unique constraint.
• Rows do not get updated using the insert commnand.

REPLACE INTO overwrites the changes. To make it clear, REPLACE options removes all rows from the table and inserts the rows from the file.

INSERT appends data to the table. If a primary or unique key violation is found, the row is deleted from the table. You can capture the deleted rows and save in an EXCEPTION table.

Read the LOAD command description in the manuals.