Fips : Dynamic File Prefetching Scheme based on File Access Patterns

Fips : 파일 접근 유형을 고려한 동적 파일 선반입 기법

  • 이윤영 (넷컴스토리지) ;
  • 김재열 (한국전자통신연구원 리눅스연구팀) ;
  • 서대화 (경북대학교 공과대학 전자전기공학부)
  • Published : 2002.08.01

Abstract

A Parallel file system is normally used to support excessive file requests from parallel applications in a cluster system, whereas prefetching is useful for improving the file system performance. This paper proposes a new prefetching method, Fips(dynamic File Prefetching Scheme based on file access patterms), that is particularly suitable for parallel scientific applications and multimedia web services in a parallel file system. The proposed prefetching method introduces a dynamic prefetching scheme to predict data blocks precisely in run-time although the file access patterns are irregular. In addition, it includes an algorithm to determine whether and when the prefetching is performed using the current available I/O bandwidth. Experimental results confirmed that the use of the proposed prefetching policy in a parallel file system produced a higher file system performance.

병렬 파일시스템은 클러스터 시스템에서 과도한 입출력 요청을 원활하게 지원하기 위해 사용되며, 특히 파일 선반입은 병렬 파일시스템의 성능을 개선하는데 유용하게 사용된다. 본 논문은 과학계산용 병렬 응용과 멀티미디어 서버 응용에서 효과적인 파일 접근 유형을 고려한 새로운 동적 파일 선반입기법인 Fips를 제안한다. 본 논문이 제안하는 동적 파일 선반입 기법인 Fips는 파일의 접근 유형을 고려하여 동적으로 선반입 할 데이타 블록을 예측하고, 다양한 접근 유형에서도 데이타 블록의 선반입을 효율을 높였다. 그리고 현재의 가용 대역폭을 고려하여 선반입 시기를 결정하므로 선반입이 시스템에 과부하로 작용하는 것을 방지하도록 하였다. 병렬 파일시스템에 Fips를 적용하여 실험한 결과 다양한 작업부하에서 제안한 선반입 기법은 우수한 성능을 보여주었다.

Keywords

References

  1. L. Breslau, P. Cao, L. Fan, G. Phillips and S. Shenker, 'Web Caching and Zipf-like Distributions : Evidence and Implications,' In Proc. of IEEE Infocom '99, pp. 126-134, March 1999 https://doi.org/10.1109/INFCOM.1999.749260
  2. Evgenia Smirni, Daniel A. Reed, 'Workload Characterization of Input/Output Intensive Parallel Applications,' Proc. of the Conference on Modelling Techniques and Tools for Computer Performance Evaluation, Springer-Verlag Lecture Notes in Computer Science, Jun. 1997, Vol. 1245, pp. 169-180 https://doi.org/10.1007/BFb0022205
  3. N. Nieuwejaar, D. Kotz, A. Purakayastha, C. S. Ellis and M. L. Best, 'File-access characteristics of parallel scientific workloads,' Parallel and Distributed Systems, IEEE Transactions on Vol. 7, Oct. 1996, pp. 1075 -1089 https://doi.org/10.1109/71.539739
  4. W. B. Ligon III and R. B. Ross, 'An Overview of the Parallel Virtual File System,' In Proc. of the 1999 Extreme Linux Workshop, June 1999
  5. Ekechi K. E. Nwokah, 'Parallel File Access On Workstation Clusters,' Ph.D. Thesis, Purdue University, West Lafayette, IN, 1999
  6. T. Cartes, 'Cooperative Caching and Prefetching in Parallel/Distributed File Systems,' Ph.D Thesis, Universitat Politecnica de Catalunya, 1997
  7. R. H. Patterson and G. A. Gibson, 'Exposing I/O concurrency with informed prefetching,' In Proc. Third International Conf. on Parallel and Distributed Information Systems, pp. 7-16, September 1994
  8. J. Cho, C. Kim, and D. Seo, 'A Parallel File System Using Dual Cache Scheme and Prefetching,' The 2000 International Conference on Parallel/Distributed Processing Techniques and Application (PDPTA2000), June, 2000
  9. J. K. Ousterhout, H. Da Costa, D. Harrison, J. A. Kunze, M. Kupfer, and J. G. Thompson, 'A Trace-Driven Analysis of the UNIX 4.2 BSD File System,' In Proc. of the 10th Symposium on Operating System Principles, pp. 15-24, December 1985
  10. M. G. Baker, J. H. Hartman, M. D. Kupper, K. W. Shirriff and J. K. Ousterhout, 'Measurements of a Distributed File System,' In Proc. of 13th ACM Symposium on Operating Systems Principles, Association for Computing Machinery SIGOPS, pp. 198-212, October 1991
  11. D. Kotz and C.S. Ellis. 'Practical prefetching techniques for multiprocessor file systems,' Journal of Distributed and Parallel Databases, 1(1):33-51, January 1993 https://doi.org/10.1007/BF01277519
  12. R.H. Patterson, G.A. Gibson, E. Ginting, D. Stodolsky, and J. Zelenka. 'Informed prefetching and caching,' In Proc. of the Fifteenth ACM Symposium on Operating Systems Principles, pp. 79-95, December 1995
  13. T. M. Madhyastha and D. A. Heed, 'Exploiting Global Input/Output Access Pattern Classification,' In Proc. of SC'97, November 1997, CD-ROM