Semantic partitioning as a basis for parallel I/O in database management systems

Modern applications such as 'video on demand' require fast reading of compl ete files, which can be supported well by file striping, Many conventional applications, however, are only interested in some part of the available re cords. In order to avoid reading attributes irrelevant to such applications , each attribute could be stored in a separate (transposed) file. Aiming at I/O parallelism, byte-oriented striping could be applied to transposed fil es. However, such a fragmentation ignores the semantics of data. This fragm entation cannot be optimized by a database management system (DBMS) because a DBMS has to perform its tasks on the basis of data semantics. For exampl e, queries must be translated into file operations using a scheme that maps a data model to a file system. However, details about files, such as the striping width, are invisible to a DBMS. Therefore, we propose to store each transposed file related to a co mposite type on a separate, independent disk drive, which means I/O paralle lism tuned to a data model. As we also aim at system reliability and data a vailability, each transposed file must be duplicated on another drive. Cons equently, a DBMS also has to guarantee correctness and completeness of the allocation of transposed fries within an array of disk drives. As a solutio n independent of the underlying data model, we propose an abstract framewor k consisting of a meta model and a set of rules. (C) 2000 Elsevier Science B.V. All rights reserved.