Data Processing

Oracle LSH can use several different methods to process data internally when you run a Program that writes data to a table. The method used depends on the setting of an attribute of the target Table instance, not the Program, though the Program type must be compatible. The way Oracle LSH handles record deletion and auditing in a table depends on the processing mode specified for the Table instance.

The data processing types available are:

• Reload. Reload processing requires a target Table instance with a primary or unique key defined. Each time you run a Load Set or a Program whose target Table instances have the Reload processing type, Oracle LSH processes all data, comparing the primary or unique key of each incoming record with the primary or unique key of existing records. The system updates records that have changed since the last load, inserts new records, and updates the refresh timestamp of all records. Reload processing is always audited. Reload processing is appropriate for use with Load Sets and with SAS Programs that write to tables.

  If you specify Full Reload processing at runtime, the system also soft-deletes all records that are no longer present in the source. If you choose Incremental Reload, the system does not delete any records, making the processing go faster.

• Transactional. Transactional processing is appropriate for use with Oracle Reports and PL/SQL Programs. The Program's source code includes explicit Insert, Update, and Delete statements. You can specify whether or not you want the data changes to be audited. If you choose to audit, the system performs soft-deletions only; deleted data remains in the database associated with an end timestamp.
• Staging. Staging processing is appropriate for internal use with any Program type. The system deletes, or appears to delete, all data from the Table instance immediately before each Program execution. You can specify whether or not you want to use auditing. If you choose to audit, the system leaves all records in the Table instance, but future executions of the same Program cannot see them. If you choose not to audit, the data is hard-deleted and cannot be reconstructed later.

See "Execution and Data Handling" in the Oracle Life Sciences Data Hub Application Developer's Guide for further information.