Restructure function overview

The Restructure function can be used to resize an area in a DEDB, add and remove areas, or change the randomizer in a DEDB. If the DEDB uses data compression, you can add, change, or remove the compression. If the DEDB uses PFX indexes or IBM native indexes, and is being restructured, you can add, change, or delete indexes. You can also perform application processing positioning changes, such as inserting a new segment at the end of a hierarchical path.

The Restructure Plan data set (created by the Prepare function) is used as input into the Restructure function. The Restructure function creates new (shadow) area data sets that are modified to match the new database definition parameters, and contains the up-to-the-minute production application data.

The shadow area data sets must be initialized before the RESTRUCTURE command is executed. For more information, see Initializing shadow area data sets.

During restructure processing, all restructured areas in the original database are copied to the target areas in the shadow database. Online activity (updates) are then captured and applied until the post-processing activities occur.

Fast Path Online Restructure/EP does not tolerate C (critical), E (error), or F (fatal), pointer errors during restructure processing. The Restructure function terminates if an E-level or a C-level pointer error is encountered. As a result, a message is issued, indicating the pointer error, along with a fatal error (RC=16).

This section contains the following topics:

• Area Copy task
• Change Capture task
• Change Apply task
• Post-processing task