Competitive parallel disk prefetching and buffer management

被引:1
|
作者
Barve, R [1 ]
Kallahalla, M
Varman, PJ
Vitter, JS
机构
[1] Duke Univ, Dept Comp Sci, Durham, NC 27708 USA
[2] Rice Univ, Dept Elect & Comp Engn, Houston, TX 77251 USA
来源
JOURNAL OF ALGORITHMS-COGNITION INFORMATICS AND LOGIC | 2000年 / 36卷 / 02期
基金
美国国家科学基金会;
关键词
D O I
10.1006/jagm.2000.1089
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
We provide a competitive analysis framework for online prefetching and buffer management algorithms in parallel I/O systems, using a read-once model of black references. This has widespread applicability to key I/O-bound applications such as external merging and concurrent playback of multiple video streams. Two realistic lookahead models, global lookahead and local Lookahead, are defined. Algorithms NOM and GREED, based on these two forms of lookahead are analyzed for shared buffer and distributed buffer configurations, both of which occur frequently in existing systems. An important aspect of our work is that we show how to implement both of the models of lookahead in practice using the simple techniques of forecasting and flushing. Given a D-disk parallel I/O system and a globally shared I/O buffer that can hold up to M disk blocks, we derive a lower bound of Omega(root D) on the competitive ratio of any deterministic online prefetching algorithm with O(M) lookahead. NOM is shown to match the lower bound using global M-block lookahead. In contrast, using only local lookahead results in an Omega(D) competitive ratio. When the buffer is distributed into D portions of M/D blocks each, the algorithm GREED based on local lookahead is shown to be optimal, and NOM is within a constant factor of optimal. Thus we provide a theoretical basis for the intuition that global lookahead is more valuable for prefetching in the case of a shared buffer configuration, whereas it is enough to provide local lookahead in the case of a distributed configuration. Finally, we analyze the performance of these algorithms for reference strings generated by a uniformly-random stochastic process and we show that they achieve the minimal expected number of I/Os. These results also give bounds on the worst-case expected performance of algorithms which employ randomization in the data layout. (C) 2000 Academic Press.
引用
收藏
页码:152 / 181
页数:30
相关论文
共 50 条
  • [41] A space-efficient on-disk prefetching algorithm
    Park, Junseok
    Koh, Kern
    ICCSA 2007: PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE AND APPLICATIONS, 2007, : 265 - +
  • [42] Dynamic Buffer Management in Massively Parallel Systems: The Power of Randomness
    Pham, Minh
    Yuan, Yongke
    Li, Hao
    Mou, Chengcheng
    Tu, Yicheng
    Xu, Zichen
    Meng, Jinghan
    ACM TRANSACTIONS ON PARALLEL COMPUTING, 2025, 12 (01)
  • [43] Analysis of simple randomized buffer management for parallel I/O
    Kallahalla, M
    Varman, PJ
    INFORMATION PROCESSING LETTERS, 2004, 90 (01) : 47 - 52
  • [44] Accelerating MCMC via Parallel Predictive Prefetching
    Angelino, Elaine
    Kohler, Eddie
    Waterland, Amos
    Seltzer, Margo
    Adams, Ryan P.
    UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, 2014, : 22 - 31
  • [45] Near-optimal parallel prefetching and caching
    Kimbrel, T
    Karlin, AR
    SIAM JOURNAL ON COMPUTING, 2000, 29 (04) : 1051 - 1082
  • [46] Integrated prefetching and caching in single and parallel systems
    Albers, S
    Büttner, M
    INFORMATION AND COMPUTATION, 2005, 198 (01) : 24 - 39
  • [47] Near-optimal parallel prefetching and caching
    Kimbrel, T
    Karlin, AR
    37TH ANNUAL SYMPOSIUM ON FOUNDATIONS OF COMPUTER SCIENCE, PROCEEDINGS, 1996, : 540 - 549
  • [48] Adaptive file prefetching in parallel file system
    Hwang-Bo, JH
    Lim, JD
    Seo, DW
    PDPTA'2001: PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED PROCESSING TECHNIQUES AND APPLICATIONS, 2001, : 1919 - 1924
  • [49] Strategy for dynamic data prefetching and buffer allocation in VoD system
    Jiang, Liquan
    Zhu, Yangyong
    Chen, Lianggang
    Jisuanji Gongcheng/Computer Engineering, 1999, 25 (11): : 21 - 23
  • [50] Adaptive prefetching using global history buffer in multicore processors
    Mahmood Naderan-Tahan
    Hamid Sarbazi-Azad
    The Journal of Supercomputing, 2014, 68 : 1302 - 1320