You can preempt spooling preceding a table copy when using Teradata PT API. The Teradata Database writes data to temporary disk space (spool space) where it is held until the select is complete. This action is known as spooling. Although spooling occurs by default to prevent data inconsistencies, it also adds to the time required to copy database objects. To improve performance for a specific table copy in jobs that use Teradata PT API, disable spooling for the table.
- Locks are maintained during the entire copy process.
- Because the spooling phase detects data conversion errors, eliminating spooling when copying a table can introduce conversion errors. If this happens, correct the errors and restart the job.
The export_without_spool element is optional and enabled by default. This element can only be specified in the XML, but not as a parameter on the command-line interface. The element is only available for Teradata PT API jobs, and only applies to Teradata Database 13.10 and later. If you specify true for the value of the export_without_spool element in the XML, and the source or target is earlier than Teradata Database 13.10, or if the job uses a different utility, the element is ignored. This element is under the table element in the XML file. The following table lists valid values for the export_without_spool element.
|false||Data Mover writes to spool space when copying a table.|
|true||Data Mover copies a table without writing to spool space.
This is the default value.
|unspecified||Equivalent to omitting the export_without_spool element in the XML. Data Mover does not use spooling when exporting data from the source Teradata Database in a table copy job that uses the Teradata PT API.|
<database selection="unselected"> <name>srcDatabase</name> <table selection="included"> <name>Employees</name> <export_without_spool>true</export_without_spool> </table> </database>