Tight bounds for prefetching and buffer management algorithms for parallel I/O systems

被引:9
|
作者
Varman, PJ
Verma, RM
机构
[1] Rice Univ, Dept Elect & Comp Engn, Houston, TX 77005 USA
[2] Univ Houston, Dept Comp Sci, Houston, TX 77204 USA
基金
美国国家科学基金会;
关键词
parallel I/O; algorithms; prefetching; caching; buffer management; competitive ratio; multiple-disk systems; external memory;
D O I
10.1109/71.819948
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
The I/O performance of applications in multiple-disk systems can be improved by overlapping disk accesses. This requires the use of appropriate prefetching and buffer management algorithms that ensure the most useful blocks are accessed and retained in the buffer. in this paper, we answer several fundamental questions on prefetching and buffer management for distributed-buffer parallel I/O systems. First, we derive and prove the optimality of an algorithm, P-min, that minimizes the number of parallel I/Os. Second, we analyze P-con, an algorithm that always matches its replacement decisions with those of the well-known demand-paged MIN algorithm. We show that P-con can become fully sequential in the worst case. Third, we investigate the behavior of on-line algorithms for multiple-disk prefetching and buffer management. We define and analyze P-Iru, a parallel version of the traditional LRU buffer management algorithm. Unexpectedly, we find that the competitive ratio of P-Iru is independent of the number of disks. Finally, we present the practical performance of these algorithms on randomly generated reference strings. These results confirm the conclusions derived from the analysis on worst case inputs.
引用
收藏
页码:1262 / 1275
页数:14
相关论文
共 50 条
  • [11] An energy-oriented evaluation of buffer cache algorithms using parallel I/O workloads
    Yue, Jianhui
    Zhu, Yifeng
    Cai, Zhao
    IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 2008, 19 (11) : 1565 - 1578
  • [12] Energy-Aware Prefetching for Parallel Disk Systems Algorithms, Models, and Evaluation
    Manzanres, Adam
    Ruan, Xiaojun
    Yin, Shu
    Nijim, Mais
    Luo, Wei
    Qin, Xiao
    2009 8TH IEEE INTERNATIONAL SYMPOSIUM ON NETWORK COMPUTING AND APPLICATIONS, 2009, : 90 - +
  • [13] Asymptotically tight bounds for performing BMMC permutations on parallel disk systems
    Cormen, TH
    Sundquist, T
    Wisniewski, LF
    SIAM JOURNAL ON COMPUTING, 1998, 28 (01) : 105 - 136
  • [14] BUFFER MANAGEMENT ALGORITHMS FOR RELATIONAL DATABASE-MANAGEMENT SYSTEMS
    HULL, MEC
    CAI, FF
    BELL, DA
    INFORMATION AND SOFTWARE TECHNOLOGY, 1988, 30 (02) : 66 - 80
  • [15] Combating I-O bottleneck using prefetching: model, algorithms, and ramifications
    Akshat Verma
    Sandeep Sen
    The Journal of Supercomputing, 2008, 45 : 205 - 235
  • [16] Combating I-O bottleneck using prefetching: model, algorithms, and ramifications
    Verma, Akshat
    Sen, Sandeep
    JOURNAL OF SUPERCOMPUTING, 2008, 45 (02): : 205 - 235
  • [17] Hiding I/O Latency with Pre-execution Prefetching for Parallel Applications
    Chen, Yong
    Byna, Surendra
    Sun, Xian-He
    Thakur, Rajeev
    Gropp, William
    INTERNATIONAL CONFERENCE FOR HIGH PERFORMANCE COMPUTING, NETWORKING, STORAGE AND ANALYSIS, 2008, : 242 - +
  • [18] Election in Fully Anonymous Shared Memory Systems: Tight Space Bounds and Algorithms
    Imbs, Damien
    Raynal, Michel
    Taubenfeld, Gadi
    STRUCTURAL INFORMATION AND COMMUNICATION COMPLEXITY, SIROCCO 2022, 2022, 13298 : 174 - 190
  • [19] Semantic partitioning as a basis for parallel I/O in database management systems
    Bakker, JA
    PARALLEL COMPUTING, 2000, 26 (11) : 1491 - 1513
  • [20] Application and user-specific data prefetching and parallel read algorithms for distributed file systems
    Nalajala, Anusha
    Ragunathan, T.
    Naha, Ranesh
    Battula, Sudheer Kumar
    CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS, 2024, 27 (03): : 3593 - 3613