期刊文献+

MDBS中一种改进的并发事务调度算法

An Improved Scheduling Algorithm of Concurrent Transactions in MDBS
在线阅读 下载PDF
导出
摘要 MDBS中并发事务的调度策略必须满足可串行化准则,论文主要分析以事务提交图为中心的调度算法TM2,虽然保证了全局事务提交顺序的可串行化,但在提交时才进行冲突检测方式存在缺点,提出改进后的事务调度算法TM3不仅保证了全局事务的可串行化和防止了全局死锁的发生,还提高了全局事务执行的并发度。最后通过实验在数据库加速引擎中进行模拟,对比了两种调度算法的性能。 The scheduling strategies of concurrent transactions must satisfy the criteria of serializability in Multidatabase System. This paper mainly discusses the transaction commit graph algorithm TM2 which ensures global serializability of the committing sequences, but the method of conflict testing until the committing operation exists some disadvantages. The improved algorithm TM3 firstly judges the conflict and then tests the committing loop in the commit graph, which not only ensures the conflict serializability of global transactions and keeps from global deadlock, but also increases the executing concurrent degree of global transaction.
出处 《计算机工程与应用》 CSCD 北大核心 2005年第31期11-13,40,共4页 Computer Engineering and Applications
基金 国家信息产业部电子发展基金项目
关键词 事务提交图 全局事务 全局可串行化 transaction commit graph,global transaction,global serializability
  • 相关文献

参考文献8

  • 1廖国琼,李陶深.分布式工程数据库系统中事务提交机制的研究[J].计算机辅助设计与图形学学报,2001,13(4):357-361. 被引量:6
  • 2廖国琼,刘云生.分布式主动实时数据库基于多优先级队列的优先级分派[J].小型微型计算机系统,2003,24(5):880-883. 被引量:1
  • 3Kamel Barkaoui ,Rabah Benamara. On concurrency control in multidatabase systems with an extended transaction model[J].The Journal of Supercomputing,2003 ;24(2).
  • 4卢炎生,潘怡,赵晓松,毛澄映.基于时间戳的并行实时事务调度策略[J].小型微型计算机系统,2005,26(4):624-627. 被引量:3
  • 5肖卫军,卢正鼎,李兵,李瑞轩.一种多数据库事务模型[J].小型微型计算机系统,2003,24(12):2224-2227. 被引量:6
  • 6Yuri Breitbart,Avi Silberschatz,Glenn R Thompson. Reliable Transaction Management in a Multidatabase System[C].In:Proceedings of the 1990 ACM SIGMOD international conference on Management ofdata,Atlantic City,New Jersey,1990-05.
  • 7Sharad Mehrotra,Rajeev Rastogi,Yuri Breitbart et al.Efficient global transaction management in multidatabase systems[R].Technical Report TR-92-11 ,Department of Computer Sciences,University of Texas at Austin, 1992.
  • 8S Mehrotra,R Rastogi,Y Breitbart et al. Ensuring transaction atomi city in multidatabase systems[C].In:Proceedings of the Eleventh ACM SIGACT-SIGMOD-SIGART Symposium on Principles of DatabaseSystems,San Diego, 1992.

二级参考文献15

  • 1郑振楣,分布式数据库,1998年
  • 2Ho Dong Yoo,Inform Sci,1995年,82卷,1/2期,75页
  • 3Dewitt D, Gray J. Parallel database systems: The future of high performance database processing [J]. Communications of ACM, 1992, 35(6):85-98.
  • 4Stonebraker M. The case for shared nothing [J]. Database Engineering, 1986,9(1): 17-24.
  • 5Dandamudi S P, Chow C Y. Performance of transaction scheduling policies for parallel database systems [C]. Proc.11^th IEEE Conf. Distributed Computing System, 1991.
  • 6Abbott R, Carcia-Molina H. Scheduling real-time transactions:A performance evaluation [J]. ACM Trans Database Systems,1992, 17(3): 513-560.
  • 7Ulusoy O, Analysis of concurrency control protocols for realtime database systems [J]. Information Science, 1998, 111:19-47.
  • 8Takkar S, Dandamudi S P. An adpative scheduling policy for real-time parallel database systems [C]. Third International Conference on Massively Parellel Computing Systems. (MPCS)1998. Colorado,USA April6-9,1998.
  • 9Averill M. Law, W. David kelton. Simulation Modeling and Analysis (3thd Edition) [M]. Beijing: Tsinghua University Press, 2000.
  • 10乔香珍.并行计算时间模型和并行机系统性能[J].计算机学报,1998,21(5):413-418. 被引量:6

共引文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部