Obtaining statistics for database tuning
You can request HSSR Engine to create optional Database Tuning Statistics reports during the sequential processing or unloading of a database. This functionality is provided as the Database Tuning Statistics function.
By activating the Database Tuning Statistics function, you can generate the following statistics reports:
- DB Statistics report
- Randomizing Statistics report
You can use these reports for the following purposes:
- Periodically monitor the need for reorganization of HDAM, HIDAM, and HISAM databases.
- Monitor the effectiveness of HDAM randomizing parameters (such as the size of the root addressable area, the number of RAPs, the block/CI size, or the bytes limit).
The Database Tuning Statistics function provides the following additional capabilities to aid your database tuning:
- Creating the optional DB Record Length Distribution report with your own definition of the database record length ranges.
- Creating the HSSRLOUT data set to record the length of each database record.
- Using the FABHLDBR utility to print long database records with the HSSRLOUT data set as an input.
Tip: You can use the FABHEXTR exit routine of the FABHURG1 utility to
create a small database extract that can be used to perform database tuning experiments. Creation of
a smaller extract database can be useful when the database administrator intends to tune the
randomizing parameters of a large HDAM database through an iterative try-and-see process. For
information about the FABHEXTR exit routine, see Creating a database extract for tuning experiments.