Loss of z108 table with rerun of 18-19 Upgrade Express; need to back up tables

• Article Type: General
• Product: Aleph
• Product Version: 18.01

Description:
The /upgrade_express_1801_1901/source/1013 step moves the z108 from the xxx30 library to the xxx50 library and deletes the z108 in the xxx30 library. If it finds an existing z108 in the xxx50 library, it deletes it.

The problem is that re-running the 1013 step will cause the z108 table to be lost. (The proc will delete it in the xxx50 library and, of course, it was deleted from the xxx30 library by the first run, so now it is gone.)

Also, the /upgrade_express_1801_1901/source/1012 step adds the "0001" sequence number to the CNO field. It doesn't check if the "0001" already exists, and the site ends up with a sequence of "00010001".

Resolution:
The Upgrade kit is not meant to be run and then re-run on the same data and configuration files.

Customers should run the kit once, check the log file, and if there is a need to run some steps again, the configuration and data should be brought again from the source version.

Since the Oracle data might be very large, the kit includes a utility for keeping a backup copy of the Oracle tables which the kit modifies ("6. Backup/Restore Oracle Tables to be Upgraded ").

If you look at the beginning of the script upgrade_express_1801_1901/util/upgrade_util_ora, you can see that Z00 (in the CRS libraries) and Z108 are kept as a backup copy.

Therefore, before re-running steps 1012 and 1013, restore the xxx30 z00 and the xxx50 z108, and then the log files should be clean of errors.

In regard to the "00010001", see also KB 16384-15667.

• Article last edited: 10/8/2013