Parallel Database Management
Multiprocessor database management
Parallel database management refers to the management of data in a multiprocessor computer and is done by a parallel database system, i.e., a full-fledge DBMS implemented on a multiprocessor computer. The basic principle employed by parallel DBMS is to partition the data across multiprocessor nodes, in order to increase performance through parallelism and availability through replication. This enables supporting very large databases with very high query or transaction loads.
Parallel database systems can exploit distributed database techniques. In particular, database partitioning is somewhat similar to database fragmentation. Essentially, the solutions for transaction management, i.e., distributed concurrency control, reliability, atomicity, and replication, can be reused. However, the critical issues for parallel database systems are data placement, parallel query processing (including optimization and execution) and load...
- 3.Özsu MT, Valduriez P. Principles of distributed database systems. 3rd ed. New York: Prentice Hall; 2009.Google Scholar
- 4.Valduriez P. Parallel database systems: the case for shared-something. In: Proceedings of the 9th International Conference on Data Engineering; 1993. p. 460–5.Google Scholar
- 5.Valduriez P, Pacitti E. Parallel Database Systems. In: Hammer J, Scheider M. Handbook of database technology. Boca Raton: CRC Press; 2007.Google Scholar