Siebel Database Upgrade Guide for DB2 UDB for z/OS > Generating the Siebel Upgrade Files > Preparing and Executing the Non-Additive Schema and JCL Files on the z/OS Host >

Creating the Data Migration Indexes

Perform the following procedure to create temporary tables and indexes for the data migration scripts and to rebuild indexes.

To create data migration indexes

1. On the Staging Database Additive Schema Menu, select option 5: Build Data Migration Index Statements, and press Enter.

   Messages are displayed on screen indicating that data migration CEATE INDEX and DROP INDEX statements are being generated. The messages you receive depends on your upgrade path.

2. On the Staging Database Non-Additive Schema Menu, select option 6: Create Data Migration Temp Table/Indexes, and press Enter.

   This job creates COMMON temp tables and indexes used by upgrade data migration SQL scripts.

3. Submit the JCL in one of the following datasets:
   • For Siebel Industry Application (SIA) upgrades, use DSNHLQ.SIEBEL.INSTALL.JCL(DMXSIA)
   • For Siebel Business Application upgrades, use DSNHLQ.SIEBEL.INSTALL.JCL(DMXHOR).
4. Verify that the job ran successfully:
   1. Review the output in SDSF or another job output facility very carefully. Verify that the RC=0, RC=4,or RC=8 and that the JCLTEST return code=FLUSH. If you do not see the FLUSH return code, you can verify the condition codes by searching for the condition code IEF206I.
   2. Verify that additional log entries have been loaded to the TMP_SBLLOG_TAR table on the staging database.
5. After submitting the job, enter cancel on the command line or press PF3 to save changes.
6. On the Staging Database Non-Additive Schema Menu, select option 7: Generate Index Rebuilds, and press Enter.

   A message appears, asking you to confirm that the DSNHLQ.SIEBEL.PROC(ISPBAT) (ISPF batch proc) is correctly configured.

   You can choose to generate index rebuilds in either TSO foreground mode or batch mode. It is recommended that you perform the procedure in batch mode.

   In foreground mode, messages are displayed on the screen as the JCL-Prep progresses. This mode will lock up your session until the option is complete. Each option can take an extended period of time (more than thirty minutes), depending on the user's dispatching priority.

   NOTE:  It is recommended that you perform the procedure in batch mode. Before doing so, make sure you modify the ISPF batch procedure, dsnhlq.siebel.proc(ispbat), to your installation standards. If you perform the procedure in foreground mode, make sure your logon region size is at least 7092.

7. Enter Y to confirm that the ISPF batch proc is correctly configured, and press Enter.
8. When the following message appears, specify the number of indexes to be included in each rebuild job, and press Enter:

   NUMBER OF INDEXES PER REBUILD JOB.

   The maximum number of indexes that can be included in a job is 10. It is recommended that you specify 3.

   CAUTION:  Consider your objective before choosing a maximum number of indexes for each job. Increasing this number results in fewer jobs but requires more memory and sort work. Reducing this number results in more jobs, which reduces resource requirements but causes fewer indexes to be built in parallel.

9. Press Enter, and you are placed in Edit mode in the DSNHLQ.SIEBEL.INSTALL(SBLJCL24) dataset.
10. Submit the job and verify that the job ran successfully.

    Review the output in SDSF or another job output facility very carefully. Verify that the RC=0 or RC=4 and that the JCLTEST return code=FLUSH. If you do not see the FLUSH return code, you can verify the condition codes by searching for the condition code IEF206I.

11. Press PF3 twice to return to the Staging Database File Generation Menu.