期刊文献+

Lustre文件系统中范围锁机制的应用研究 被引量:5

Application and research of extent lock in the Lustre
下载PDF
导出
摘要 分布式锁机制是分布式文件系统的重要机制,Lustre是一种高性能的分布式文件系统,针对现有的范围锁机制,在处理多客户端大量并发的密集型共享文件访问时,存在操作延迟、文件数据一致性管理复杂的问题,为了解决这类问题采用了可伸的锁范围划定、对范围锁赋予周期、范围锁版本控制相结合的改进机制。实验结果表明,采用改进后的范围锁机制,不仅保证了文件数据的一致性,一定程度上降低了并发文件访问延迟,从而保证了分布式文件系统的I/O性能。 Distributed lock-management is the important role of the distributed filesystem.Lustre is a High-performance distributed filesystem.For its existing range of locking mechanisms,in dealing with multiple clients sharing a large of concurrent file access intensive,there is delay in operation,file management of complex issues of data consistency.To solve such problems can be extended by the delineation of the scope of the lock,the lock on the range given cycle,the scope of the combination lock to improve version control mechanism.The results show that,using this mechanism to effectively ensure the consistency of file data,concurrent file access latency reduction,thereby enhancing the distributed file system I/O performance.
出处 《重庆邮电大学学报(自然科学版)》 北大核心 2011年第2期242-246,共5页 Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金 重庆市信息产业发展专项资金(200811004)~~
关键词 分布式锁 数据一致性 LUSTRE 范围锁 distributed lock data consistency lustre extent lock
  • 相关文献

参考文献3

二级参考文献19

  • 1更实用的SAN存储[J].中国计算机用户,2003(6):36-36. 被引量:2
  • 2Braam P J. Lustre: A Scable, High-Performance File Systme [M]. Lustre Whitepaper Version 1.0,2002.
  • 3Davis R G. VAXcluster Principles[M]. Digital Equipment Corporation, 1993.
  • 4Thomas K. Programming Locking Applications v. 4. 3. 1 [EB/OL]. [1999-05-08]. http: ffwww. 124. ibm. com/developerworks/oss/dlm/currentbook/dlmbook_index, html.
  • 5Kistler J J, Satyanarayanan M. Disconnected Operation in the Coda File System[J]. ACM Trans on Computer Systems, 1992,10(1):3-25.
  • 6Burns R C, Rees R M, Long D D E. Semi-Preemptible Locks for a Distributed File System[C]// Proc of 2000 Int'l Performance Computing and Communication Conf, 2000.
  • 7Schmuck F, Haskin R. GPFS: A Shared-Disk File System for Large Computing Cluters[C]//Proc of Conf on File and Storage Technologies, 2002.
  • 8国防科技大学.三种SAN架构比较[EB/OL](2003-3-31)[2005-06-21].http://www.ccw.com.cn/applic/tech/htm2003/20030331_14kx7.htm.
  • 9存储网络架构的核心要素(白皮书)[EB/OL](2005-4-5)[2005-08-01].http://www.techhn.com/download/resource/Architecture.pdf.
  • 10寻根-数据存储之根本[EB/OL](2005-10-26)[2005-12-11].http://media.ccidnet.com/art/2991/20051026/358777_1.html

共引文献4

同被引文献24

  • 1张钰森,吴庆波,谭郁松.基于Lustre的大规模文件系统拥塞控制算法的研究[J].计算机研究与发展,2011,48(S1):189-193. 被引量:4
  • 2韩德志,刘昊.智能存储及管理技术探析[J].计算机应用研究,2005,22(2):5-7. 被引量:12
  • 3聂刚,卿秀华.基于对象存储的Lustre文件系统的研究[J].信息技术,2007,31(9):16-18. 被引量:2
  • 4Gifford D K,Jouvelot P.Semantic file systems[J].ACM Symposium on Operating Systems Principles,1991.
  • 5Guttman A.R.-Trees:A dynamic index structure for spatial searching[J].Proc of the 1984 ACM SIGMOD Int' I Confon the Management of Data. Boston, 1984.
  • 6Bloom,BH.Space/time trade-offs in hash coding with allow-able errors[J].Communications of the ACM,1970.
  • 7Gifford D K,Jouvelot P.Semantic file systems[J].ACM Symposium on Operating Systems Principles,1991.
  • 8Guttman A.R.-Trees:A dynamic index structure for spatial searching[J].Proc of the 1984 ACM SIGMOD Int'l Confon the Management of Data. Boston, 1984.
  • 9Bloom,BH.Space/time trade-offs in hash coding with allow-able errors[J].Communications of the ACM,1970.
  • 10高俊,夏运均,游雄,等.虚拟现实在地形环境仿真中的应用[M].北京:电子工业出版社,1999.

引证文献5

二级引证文献3

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