Load Utility in DB2

The load utility is capable of efficiently moving large quantities of data into tables.
The utility can handle all data types, including large objects (LOBs) and user-defined types (UDTs).
The load utility is faster than the import utility, because it writes formatted pages directly into the database, while the import utility performs SQL INSERTs.
The load utility does not fire triggers, and does not perform referential or table constraints checking (other than validating the uniqueness of the indexes).

 

db2 "load from data.ixf of ixf insert into abc.data"

 

Syntax

 

Load from filename | pipename | device Of filetype

Lobs from lobpath modified by filetype_mod method load_method

[Savecount n] [Rowcount n] [Warningcount n] [Messages message_file]

Insert | Replace | Restart | Terminate into table_name [(column {, column})]

 

-- Difference between Load and Import in DB2

 

 

IMPORT

 

1. Inserts one record at a time.

2. Before a row is inserted, it checks for any constraint violations. Only if the inserted row complies with the rules, the row is inserted.

3. After a row in inserted , the change is logged. (Rollback is possible)

4. Table may or may not exist.

 

LOAD

 

1. Reads all the records from the external file and builds data pages out of it.

2. All the built pages are directly written to the database in a single shot.

3. Logs the entire operation and not on individual record insert (Cannot be undone/rolled back)

4. Table must exist.

5. Checks for constraint violations and duplicate rows after data pages are inserted. Any rows not complying with the rules are deleted post insert.

6. Indexes are rebuilt after insert.