The Audit Trail File
Audit log records are inserted in the audit tables you define. The base product contains a single such table (called CI_AUDIT ). However, the base audit insert program is designed to allow you to use multiple audit tables.
If you want to segregate audit information into multiple tables, you must create these tables. Use the following guidelines when creating new audit tables (that use the base delivered audit insert program):
- The new audit tables must look identical to the base table (CI_AUDIT).
- The new tables must be prefixed with CM (e.g., CM_AUDIT_A, CM_AUDIT_B, etc.). An appropriate java table name must be defined and an *_impl.java program for that java table name must be defined.
- The name of the new table must be referenced on the various tables whose changes should be logged in the new table.
Note: It's important to note if you use your own tables (as opposed
to using the base package table called CI_AUDIT), the SQL used to insert
and access audit trail records in the base delivered audit program
is dynamic. Otherwise, if the base package's table is used, the SQL
is static.