The Usasge and Performance of Object Databases Compared with ORM Tools in a Java Environment.
Mikael Kopteff · 2008
Object databases have been almost entirely forgotten in the mainstream software development world. The huge interest in object databases during the 1990’s has changed to ignorance even though objectoriented programming languages like Java and C# continue to thrive. During the last couple of years many object-relational mapping tools have emerged to solve the notorious object-relational impedance mismatch between relational databases and object-oriented languages. The widespread popularity of object-relational mapping (ORM) tools still raises the question of using an object database instead of a redundant data mapping tool and persisting data in its natural form – as objects. The goal of this work is to find out if object databases can be used as persistent storage in applications instead of ORM tools and relational databases. Two popular tools from both camps were selected as reference implementations for the study. The features, usage and performance of these tools were then studied from a software development point of view. Features of both tools were listed and compared. Usage of the tools was studied by comparing the query languages and the actual code used to access the database. Performances of the two reference implementations were tested by running a subset of the de-facto object database benchmark OO7. These three factors provide sufficient information if object databases can replace ORM tools and relational databases as the persistence layer in modern software development.