| This book is aimed at helping Oracle DBAs understand and use PeopleSoft technology. For the typical DBA, the introduction to PeopleSoft is likely to include some surprises, not all of them agreeable. Many—if not most—DBAs have to deal with many different databases, usually supporting different applications. Often they will want to be able to administer all databases in a standard fashion. However, this is not always possible with a PeopleSoft system.
Most surprising to Oracle DBAs may be what is missing. In a vanilla PeopleSoft database, there is only minimal use of Oracle-specific features and Oracle-specific SQL constructions. There are no referential constraints. Very few optimizer hints are used, and only where there is no alternative. All PeopleSoft processes connect to the one database schema that contains all the database objects, so security is maintained by the application, not the database. Oracle sequences are never used; instead, sequence numbers are generated using ordinary tables.
In order to avoid the use of platform-specific SQL constructions, most of the delivered SQL conforms to a lowest common denominator subset of SQL accepted by the supported RDBMS platforms (Oracle, Microsoft SQL Server, DB2, Sybase, and Informix). The data model is kept uniform across all platforms, although there are variations in the column data types between platforms. There can be some differences in the indexing between platforms. There is some capability for different code on different platforms in PeopleSoft, but its use in the delivered product is kept to an absolute minimum. |
