DB2 Manager User Guide 81
StorageTek Proprietary
Functions
The table migration utility performs the following functions:
• It processes and validates a series of parameters from a SYSIN file, and
an object selection parameter from file SQLIN which will govern execution
of the utility. A parameter specification and validation report is written to
the SYSPRINT dataset.
• It sequentially processes all unmigrated rows in the DB2 table identified in
the SQLIN parameters and selects all rows which satisfy the migration
criteria.
• Each selected row is passed to a writer task in the DB2 Manager control
region for storage as a component of an object in the Archive Manager
database specified in the SYSIN parameter. The number of rows per
object is controlled via the OBJSIZE parameter in the ENVCNTL member
of the DB2 Manager parameter library.
• On successful storage in the Archive Manager database, the migrated row
is replaced by an 18-byte “archive stub”, identifying that the row has been
migrated and recording its location in the Archive Manager database.
Index entries for the migrated row will be unaltered, and will continue to
exist as they had prior to row migration. The archive stub is stored in the
table’s archive partition (created during enabling of the table for migration
processing).
• The utility will commit updates to the Archive Manager database and the
DB2 table at the frequency specified in the COMMIT execution parameter.
Following successful completion of commit processing, access to all rows
migrated since the last commit point (or since the start of utility processing
for the first commit request) will be made via the DB2 Manager control
region.
• When all rows have been successfully migrated, the utility will produce a
processing report indicating successful completion and providing totals of
the number of rows processed and the number of rows migrated during
utility execution.
• If the utility encounters a serious error which prevents continued
processing, it will terminate with a condition code 12 after producing an
error report identifying the cause of the error and a processing report
indicating unsuccessful completion and giving totals of the number of rows
processed and the number of rows successfully migrated up to the last
successful commit point.