DocumentCode :
413038
Title :
Scheduling of query execution plans in symmetric multiprocessor database systems
Author :
Wu, Jun ; Chen, Jian-Jia ; Hsueh, Chih-Wen ; Kuo, Tei-Wei
Author_Institution :
Dept. of Comput. Sci. & Info. Eng., Nat. Chung Cheng Univ., Chiayi, Taiwan
fYear :
2004
fDate :
26-30 April 2004
Firstpage :
2
Abstract :
Summary form only given. While excellent research results have been proposed for query optimization, little work has been done for the scheduling of query execution plans. We target the optimization problem of the schedule lengths of query execution plans in a symmetric multiprocessor (SMP) database system. We show the NP-hardness of the optimization problem. A critical-path-based algorithm is proposed to minimize the schedule length of a collection of query execution plans. When each query execution plan has a tree or a directed acyclic graph structure, we show that the approximation ratios of our proposed algorithm are 2 and (3 - 2/M), respectively, where M is the number of processors in the system. When the proposed algorithm is adopted for online usages, the competitive ratio of the algorithm is proven being (3 $2/M). The proposed algorithm is optimal when there is a sufficient number of processors. The performance of the proposed algorithm is evaluated based on the TPC-C benchmark.
Keywords :
computational complexity; database machines; distributed algorithms; multiprocessing systems; processor scheduling; query processing; NP-hard problem; TPC-C benchmark; algorithm performance evaluation; critical-path-based algorithm; directed acyclic graph; processor scheduling; query execution plan; query optimization; symmetric multiprocessor database system; Approximation algorithms; Computer science; Councils; Data engineering; Database systems; Load management; Processor scheduling; Query processing; Scheduling algorithm; Tree graphs;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel and Distributed Processing Symposium, 2004. Proceedings. 18th International
Print_ISBN :
0-7695-2132-0
Type :
conf
DOI :
10.1109/IPDPS.2004.1302900
Filename :
1302900
Link To Document :
بازگشت