Performance Study of the Index-based Parallel Join

The index file has been used a access database records effectively. The join operation in a relational database system requires a large execution time, especially in the case of handling large size tables. If the indexes are available on the joining attributes for both relations involved in the join and the join selectivity is relatively small, we can improve the execution time of the join operation. In this paper. we investigate the performance trade-offs of parallel index-based join algorithms where different indexing schemes are used. We also present a comparison of our index-based parallel join algorithms with the hash-based parallel join algorithm.