Distributing a database for parallel processing is NP-hard
暂无分享,去创建一个
In a database the response time to a query can be reduced if certain concurrent operations are provided. In order to maximize the degree of concurrency, data has to be carefully allocated. The complexities of the following two problems are studied in this paper: one is to allocate a database onto a multiple-disk system which maximizes the disk access concurrency; the other one is to allocate a database over a network which minimize the number of nodes being used and a complete parallel searching is possible for a set of queries. Both are shown to be NP-hard (computational intractable) problems.
[1] B. Srinivasan,et al. Algorithms to Distribute a Database for Parallel Searching , 1981, IEEE Transactions on Software Engineering.
[2] John S. Sobolewski,et al. Disk allocation for Cartesian product files on multiple-disk systems , 1982, TODS.
[3] Sakti P. Ghosh. Distributing a Data Base with Logical Associations on a Computer Network for Parallel Searching , 1976, IEEE Transactions on Software Engineering.
[4] David S. Johnson,et al. The Complexity of Near-Optimal Graph Coloring , 1976, J. ACM.