期刊文献+

Active Bolck Layout:一种高性能磁盘布局机制

Active Block Layout: A High Performance Block Layout Mechanism
下载PDF
导出
摘要 The access frequency of different files in file system is dissimilar. If file system can optimize the block lay-out of these hot files which are frequently accessed,the performance of file system will be improved. This paper pre-sents a high performance block layout mechanism: Active Block Layout (ABL). ABL can record the access frequencyof every file in file system and actively optimize the block layout of these hot files by block duplicating. The duplicatedblocks can be placed in the special zone of track,which is called "Cooling Zone". ABL can automatically determine theplacing position and the copy count of the blocks which need to be duplicated. In order to reduce the overhead of blockduplication,this paper also presents a mechanism which uses the potential disk bandwidth to realize the block duplica-tion,and does not obviously degrade the performance of file system. The access frequency of different files in file system is dissimilar. If file system can optimize the block layout of these hot files which are frequently accessed,the performance of file system will be improved. This paper presents a high performance block layout mechanism: Active Block Layout (ABL). ABL can record the access frequency of every file in file system and actively optimize the block layout of these hot files by block duplicating. The duplicated blocks can be placed in the special zone of track,which is called 'Cooling Zone'. ABL can automatically determine the placing position and the copy count of the blocks which need to be duplicated. In order to reduce the overhead of block duplication,this paper also presents a mechanism which uses the potential disk bandwidth to realize the block duplication,and does not obviously degrade the performance of file system.
出处 《计算机科学》 CSCD 北大核心 2003年第3期164-167,共4页 Computer Science
基金 国家95重点攻关项目
关键词 ActiveBlockLayout 高性能磁盘布局机制 磁盘存储系统 磁盘驱动器 计算机 Disk, Performance, Layout, Active, Bandwidth
  • 相关文献

参考文献9

  • 1[3]Yu Xiang,Gum B,et al. An-derson. Trading capacity for performance in a disk array. Symposium on Operating Systems Design and Implementation (San Diego, CA, 23-25 October 2000),USENIX Association, 2000. 243 ~258
  • 2[4]Ruemmler C, Wilkes J. An introduction to disk drive modeling. IEEE Computer, 1994,27 (3): 17 ~ 28
  • 3[5]Jiri S,John L,Lumb C P,et al. Track-Aligned Extents: Matching Access Patterns to Disk Drive Characteristics. In: Proc. of the FAST '02 Conf. on File and Storage Technologies, Monterey, California,Jan. 2002. 259~274
  • 4[6]Lumb C, Schindler J,Ganger G R, Riedel E, et al. D. F. Towards Higher Disk Head Utilization: Extracting Free Bandwidth from Busy Disk Drives. In: Proc. of the Fourth Symposium on Operating Systems Design and Implementation, San Diego, CA ,Oct. 2000
  • 5[7]Baker M, Hartman J, Kupfer K, Shirriff K, Ousterhout J. Measurements of a Distributed File System. ACMSymposiumon Operating SystemsPrinciples,1991. 198~212
  • 6[8]Ruemmler, Chris, Wilkes J. UNIX disk access patterns. Usenix Conference Proceedings, Jan. 1993. 405~420
  • 7[9]Ganger G R,Kaashoek M F. Embedded inodes and explicit grouping: exploiting disk bandwidth for small files. Annual USENIX Technical Conference (Anaheim,CA),Jan. 1997.1~17
  • 8[1]Chen P M, Lee E K,Gibson G A,Katz P H,Patterson D A. RAID:High-Performance, Reliable Secondary Storage. ACM Computing Surveys, 1994,26 (2): 145~ 185
  • 9[2]Gibson G. Redundant Disk Arrays: Reliable, Parallel Secondary Storage. MIT Press, 1992

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