You must define the temp-table layout in your application program to accommodate specific result sets before you attempt to populate the temp-tables with data. If a SQL statement retrieves more than one result set, you must define multiple temp-tables to be able to retrieve all the data. Therefore, the success of this approach depends to a large extent on your:

• Understanding of the specific data your foreign data source is providing you through a given stored procedure
• Ability to correctly define temp-tables

The following types of temp-tables can support result sets:

• Static — A temp-table whose schema is defined at compile time.
• Dynamic — A temp-table whose schema is defined at run time. There are two types of dynamic temp-tables: dynamic-prepared and dynamic-unprepared.

Keep in mind that you can pass handles of temp-tables that contain a mixed array. A mixed array is one in which some of the temp-table handle elements can be static while others can be dynamic. Also, note that a stored procedure supports the use of an INT64 data type in static and dynamic temp tables when the LOAD-RESULT-INTO phrase processes the procedure's result set on the RUN-STORED-PROC statement.

The following table identifies the temp-table options for which you can plan and the requirements you must fulfill for each option.