Article Preview
TopIntroduction
The database system is arguably the most important development in the field of software engineering, and the database systems are now the underlying framework of information systems (Connolly & Begg, 2005). Object-relational (O-R) database systems extend relational database systems (RDBMSs) by adding modelling primitives from object world such as support for user-defined data types, references, inheritance and polymorphism. There is significant demand for object-relational capabilities; otherwise, vendors would not have uniformly embraced the technology (Stonebraker et al., 1999).
O-R database management systems (ORDBMSs) provide a huge increase in flexibility and support for rich data types compared to the RDBMSs. They are attractive, especially for non-traditional applications, such as geographic information systems, multimedia databases, etc., for which RDBMSs have experienced their shortcomings. Unlike object-oriented database systems, ORDBMSs support all features of RDBMSs including scalability and transaction processing performance. Moreover, language support for the new features of ORDBMSs has been added to recent releases of SQL standard, namely SQL:99 and SQL:2003.
So far, little is known about the performance implications of using these features (Carey et al. 1997). BUCKY (Benchmark of Universal or Complex Kwery Ynterfaces) is a query-oriented benchmark that tests many of the key features offered by ORDBMSs and evaluates the maturity of O-R technology relative to relational technology in the same ORDBMSs (Carey et al., 1997). It presents an O-R schema and a relational equivalent thereof. The outcome of running BUCKY on an early O-R product shows that the performance of the O-R version of the examined ORDBMSs was almost twice as slow as the relational version.
One key factor to the success of ORDBMSs is its performance. The concept of ORDBMSs, as a hybrid of relational and object-oriented database management systemis very appealing, preserving the wealth of knowledge and experience that has been acquired by RDBMSs (Stonebraker et al., 1999). Users expect short time responses for their database operations, even though the complexity of data and queries. Stonebraker claims that ORDBMSs will be the next great wave in evolution journey of DBMSs (Stonebraker & Moore, 1996).
Database migration is the process of converting a source database into a target database to be handled in its new environment. Comparing the performance of source and target databases may help the user to decide whether or not they should move into their chosen database if performance is a deciding factor (Maatuk, 2009).
The aim of this paper is to investigate the maturity of O-R technology relative to relational technology. This includes the evaluation of the performance of ORDBMSs relative to RDBMSs by applying an O-R query-oriented BUCKY benchmark on Oracle 10g database management system (DBMS). The results from this study and the original BUCKY results are compared and discussed.
The remainder of this paper is organized as follows. In Section 2, an overview of related work is given. The BUCKY benchmark is briefly described in Section 3. While Section 4 explains experimental set-up, the next section describes the first results obtained from applying the queries of the benchmark on Oracle 10g DBMS. In Section 6, we discuss the results as whole and presents lessons we have learned from this study. Section 7 concludes this paper.