ENHANCING QUERY TIME USING A VOLUME-ADAPTIVE BIG DATA MODEL OF RELATIONAL DATABASES

Abstract

Big Data has been traditionally associated with distributed systems, the reason being that the volume dimension of Big Data, it appears, can be best accommodated by the continuous addition of inexpensive resources. It is within this implementation context that the non-distributed database models such as the relational database model have been faulted and departure from their usage contemplated by the database community. The atomicity, consistency, isolation, and durability (ACID) properties of the relational database model however constitute a major attraction especially for applications that process transactions. A transaction-laden application may demand a lot more of the ACID properties of a database so as to maintain data integrity while requiring that the ever-increasing volume of data is also accommodated. This means that a one-size-fits-all database as proposed by several researchers may end up as a mirage and the current trend suggests that databases be made adaptive in the areas of their weakness rather than throw the baby away with the bath. This paper appreciates that the query time is negatively impacted as data volume increases in a relational database and therefore proposes a Big Data model of the relational database that partitions a relation thereby allowing volume to grow within partitions rather than a single relation. The results of the experiments performed show that the query time is enhanced as more data is accommodated in the partitions.