The tables in the Repository support on the following product activities:
Work ID generation and maintenance
A work ID can consist of steps to perform any of the following tasks:
Create a log mark
Define a log scan
Execute generated Structured Query Language (SQL) statements
Recover dropped objects
Generate High-speed Apply JCL
Retention of batch results
The product saves the results of a work ID in the Repository. The results include both the final status of the work ID and pointers to the various files generated as the work ID runs.
The results include the following files:
Logical log files
Drop recovery output files
- DDL files (PTF BQU1717 applied)
When the product runs a work ID, it writes a row that defines an individual run of the work ID to the Work ID History Table. It also associates any generated log marks or files with this row.
To reduce the number of tables used, the Repository applies a significant amount of denormalization to its internal tables. Columns in a number of the tables use overloading depending on some other value in the row. For example, the Work ID Step Table contains different information depending on the type of step (log mark, log scan, execute SQL, or drop recovery) as indicated in the type field in the row.
One specific table in the Repository (called the Old Objects Table) provides support for overtime processing and compression dictionaries. For more information about this support or for information about updating the Repository, see REPOS UPDATE in LOGSCAN statement. See, also, the relevant topics in the Command and syntax reference.
This section contains the following topics: