Understanding Trees and PS/nVision

PS/nVision performance may suffer when trees are used, especially when the SQL statements used to retrieve data access two or more trees at once. On some database platforms, the Where criteria used with most tree joins can cause the database optimizer to select the wrong access path, making the PSTREESELECTnn table an obstacle to performance rather than an aid. This often happens when the tree uses ranges of detail values.

To address these performance issues, PS/nVision includes techniques for building SQL that implement tree criteria, and it also implements SQL that is readily understood by database optimizers, yielding better access paths with less need for index tuning and so on. Additionally, you have control over the techniques used, so you can tune the performance of individual reports.