Improving performance of database system using architectural layer.

Database approach is significant part of the information technology. Each of us feels the increasing amount of data to be stored and processed as well as strong pressure on efficiency, the processing time and correctness of the results. Index has become an essential part of the processing that increases performance and efficiency of the data access in Select statements. On the other hand, it causes performance reduction for other data manipulation language statements. In this paper, we deal with the possibility of block changes and linking data blocks, in which actual data objects are stored. Thanks to approach change - size of the blocks for data themselves as well as for index structures, we can achieve significant improvement in overall system performance. [ABSTRACT FROM AUTHOR]