Object Oriented Databases Performance Optimization
Performance Optimization: Part of the reason that schema changes are difficult in a SQL database is because the schema design greatly affects performance.
Standard SQL database design dictates that an architect first normalizes a database schema to maximize storage efficiency, and then de-normalizes the areas that require it to optimize performance. This creates an undesirable trade-off of performance versus efficiency, but it also imposes a tradeoff of either performance or efficiency versus complexity of the object model that the schema is going to support.
Despite the array of tools and methods apparently available to a database administrator for performance optimization, they tend to just be new structures (views, indexes, stored procedures, etc.) layered on top of the tabular structure.
The best ways to improve performance are eliminating unnecessary work, distributing necessary work across the most appropriate resources (e.g. using memory effectively) and maximizing parallelism.
Traditional database optimizations are really very limited when trying to approach the real capabilities of the hardware, don’t address the largest areas of resource consumption and often create new problems of their own.