6.2.2 Operations involving a backup of the execution log information table

Execution log increases every time process instance is executed, capacity of database is pressed. When you acquire backup of execution log information while HCSC server is operating, the records in execution log information table remain even after acquiring the backup.

Therefore, you can deal with depletion of disc capacity and increase in database by acquiring backup of table, by deleting execution log, and reorganizing table of database periodically. The procedure is as follows:

1. Acquire difference backup of execution table using database copy utility (pdcopy) of HiRDB.

2. Delete execution log of process instance using cscpidelete command.

3. Reorganize the table of database using database reorganization utility (pdrorg) of HiRDB.

   If you repeat addition and deletion of execution log, storage efficiency of data becomes poor and performance degrades. In order to prevent this, reorganize the table periodically using pdrorg command (Database reorganization utility). Note that, you need to re-create the index as well since index is defined in execution log information table.

   For details on reorganization of HiRDB table, see manual "HiRDB System Operation Guide" and for details on pdrorg command, see manual "HiRDB Command Reference".

Because backup and deletion intervals depend on the nature of the user application, the administrator must handle these periodic operations.

The following figure shows an example of operations involving a backup of the execution log information table and deletion of records:

Figure 6‒3: Examples of operations involving a backup of the execution log information table and deletion of records

Description

1. Acquire a backup of the execution log information table (first backup).

2. At the second backup, acquire a differential backup from the first backup.

3. Use the execution log search command to log whose date and time fall before a specified date and time, and the process instances that are in a specified status that you intend to delete (in this example, the command searches for records that are more than 50 days old).

4. After acquiring a backup, use the execution log deletion command to delete the old records (in this example, the command deletes records that are more than 50 days old).

5. Reorganize the table.

6. At the third backup, acquire a differential backup from the second backup.

7. Use the execution log search command to find log whose date and time fall before a specified date and time, and execution log of process instances that are in a specified status that you intend to delete (in this example, the command searches for records that are more than 50 days old).

8. After acquiring a backup, use the execution log deletion command to delete the old records (in this example, the command deletes records that are more than 50 days old).

9. Reorganize the table.

10. At the fourth backup, acquire a differential backup from the third backup.

11. Use the execution log search command to find the log whose date and time fall before a specified date and time, and the process instances that are in a specified status that you intend to delete (in this example, the command searches for records that are more than 50 days old)

12. After acquiring a backup, use the execution log deletion command to delete the old records (in this example, the command deletes records that are more than 50 days old).

To Page Top