Modeling the Influence of Metadata on the Storage Size of Databases Implemented with Different Database Systems

Abstract

In his paper we investigate the influence of metadata on the storage size of databases implemented with different database systems such as MS Access, SQL Server and Oracle. Several sample databases have been designed and implemented in order to investigate the influence of every metadata component (attributes, tables, relationships and constraints) on the storage size of the entire database. The experimental results show clearly that the storage size of the database metadata depends on the number of attributes or tables in database systems. When the number of attributes or tables increases from 1 to 10, the storage size increases linearly with different average slopes which depend on the database system. The highest slope was observed for Oracle, while MS Access shows the lowest one. This approach of modeling the storage size in a linear form significantly simplifies the model of metadata on the storage size prediction, which is an important and useful guideline for the design and implementation of database system.