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 25 条
  • [21] Scheduling parallel jobs with CPU and I/O resource requirements in cluster computing systems
    Abawajy, JH
    Dandamudi, SP
    [J]. PROCEEDINGS OF THE 11TH IEEE/ACM INTERNATIONAL SYMPOSIUM ON MODELING, ANALYSIS AND SIMULATION OF COMPUTER TELECOMMUNICATIONS SYSTEMS, 2003, : 336 - 343
  • [22] ExaHDF5: Delivering Efficient Parallel I/O on Exascale Computing Systems
    Suren Byna
    M. Scot Breitenfeld
    Bin Dong
    Quincey Koziol
    Elena Pourmal
    Dana Robinson
    Jerome Soumagne
    Houjun Tang
    Venkatram Vishwanath
    Richard Warren
    [J]. Journal of Computer Science and Technology, 2020, 35 : 145 - 160
  • [23] Parallel I/O scheduling in the presence of data duplication on multiprogrammed cluster computing systems
    Abawajy, J. H.
    [J]. 2006 IEEE INTERNATIONAL CONFERENCE ON COMPUTER SYSTEMS AND APPLICATIONS, VOLS 1-3, 2006, : 1117 - 1124
  • [24] Pattern-Direct and Layout-Aware Replication Scheme for Parallel I/O Systems
    Yin, Yanlong
    Li, Jibing
    He, Jun
    Sun, Xian-He
    Thakur, Rajeev
    [J]. IEEE 27TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM (IPDPS 2013), 2013, : 345 - 356
  • [25] HPIS3: Towards a High-Performance Simulator for Hybrid Parallel I/O and Storage Systems
    Feng, Bo
    Liu, Ning
    He, Shuibing
    Sun, Xian-He
    [J]. 2014 9TH PARALLEL DATA STORAGE WORKSHOP (PDSW), 2014, : 37 - 42