Performance tradeoffs for client-server query processing

作者: Michael J. Franklin , Björn Thór Jónsson , Donald Kossmann

DOI: 10.1145/233269.233328

关键词:

摘要: The construction of high-performance database systems that combine the best aspects relational and object-oriented approaches requires design client-server architectures can fully exploit client server resources in a flexible manner. two predominant paradigms for query execution are data-shipping query-shipping We first define these policies terms restrictions they place on operator site selection during optimization. then investigate performance tradeoffs between them bulk processing. While each strategy has advantages, neither one its own is efficient across wide range circumstances. describe evaluate more policy called hybrid-shipping, which execute queries at clients, servers, or any combination two. Hybrid-shipping shown to least match "pure" policies, some situations, perform better than both. implementation hybrid-shipping raises number difficult problems an initial investigation into use 2-step optimization as way addressing issues.

参考文章(24)
Tom Atwood, Roderick G. Cattell, Object Database Standard: ODMG-93 Morgan Kaufmann Publishers Inc.. ,(1993)
Michael Stonebraker, Bruce G. Lindsay, Jim Gray, Michael J. Carey, Lawrence A. Rowe, David Beech, "The Committee for Advanced DBMS Function": Third Generation Data Base System Manifesto. international conference on management of data. pp. 396- ,(1990)
Michael J. Carey, Hongjun Lu, Some experimental results on distributed join algorithms in a local network very large data bases. pp. 292- 304 ,(1985)
Rosana S. G. Lanzelotte, Patrick Valduriez, Mohamed Zaït, On the Effectiveness of Optimization Search Strategies for Parallel Execution Spaces very large data bases. pp. 493- 504 ,(1993)
Donald Kossmann, Michael J. Franklin, A study of query execution strategies for client-server database systems University of Maryland at College Park. ,(1995)
Michael J. Carey, Hongjun Lu, Load Balancing in a Locally Distributed Database System. international conference on management of data. pp. 108- 119 ,(1986)
Philippe Futtersack, David J. DeWitt, David Maier, Fernando Velez, A study of three alternative workstation server architectures for object-oriented database systems very large data bases. pp. 107- 121 ,(1990)
B. Paul Jenq, Darrell Woelk, Won Kim, Wan-Lik Lee, Query processing in distributed ORION extending database technology. ,vol. 416, pp. 169- 187 ,(1990) , 10.1007/BFB0022170
Guy M. Lohman, Lothar F. Mackert, R* Optimizer Validation and Performance Evaluation for Distributed Queries very large data bases. pp. 537- 547 ,(1994)
Y. E. Ioannidis, Younkyung Kang, Randomized algorithms for optimizing large join queries international conference on management of data. ,vol. 19, pp. 312- 321 ,(1990) , 10.1145/93597.98740