How should table data rows be loaded back to Db2 after they have been removed by a reorg utility as discards?
Rapid Reorg for Db2 for z/OS (PRR) and IBM Reorg both have the facility to write discarded records to a
dataset to effectively remove them from the table during the Reorg.
This can be done with ROW-PROCESSING SELECT or DELETE in the Rapid Reorg for Db2 for z/OS utility.
Rapid Reorg writes discarded records to either a DISCARDDN DD of your choice or to the SYSDISC DD if specified.
This process may be used as a means of archiving old data by conveniently removing rows at reorg time. These rows could
then be loaded to an archive database for alternate access or used in some other way for testing or loaded to a data warehouse.
Release: R19, R20
Component: PFL, PRR, PFU
Rapid Reorg Row Processing writes deleted or not-selected rows to the SYSDISC DD in a DB2 UNLOAD format.
Reloading data discarded by CA Rapid Reorg:
If the discard records were written by Rapid Reorg, a Fast Load utility would use INPUT-FORMAT UNLOAD to load these rows from the discard dataset.
Reloading data discarded by IBM Reorg:
If the discard records were written by IBM Reorg a Fast Load utility would use INPUT-FORMAT SEQ to load these rows from the
discard dataset specifying definitions of all the columns accordingly.
Generate the correct load cards:
Fast Unload for Db2 for z/OS has the facility to produce load cards with a LOAD-CONTROL FASTLOAD ONLY with
OUTPUT-FORMAT LOAD and a SYSCTL01 DD allocated to provide the correct load control cards.
LOAD-CONTROL ONLY does not unload any data. It just writes the load cards into the CTLDDN DD. These can then be copied into the Load JCL.
Specific information can be found in the product User Guides for Rapid Reorg for Db2 for z/OS, Fast Load for Db2 for z/OS and IBM REORG.