BF-Tree: Approximate Tree Indexing

被引:42
作者
Athanassoulis, Manos [1 ]
Ailamaki, Anastasia [1 ]
机构
[1] Ecole Polytech Fed Lausanne, Lausanne, VD, Switzerland
来源
PROCEEDINGS OF THE VLDB ENDOWMENT | 2014年 / 7卷 / 14期
关键词
D O I
10.14778/2733085.2733094
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The increasing volume of time-based generated data and the shift in storage technologies suggest that we might need to reconsider indexing. Several workloads - like social and service monitoring - often include attributes with implicit clustering because of their time-dependent nature. In addition, solid state disks (SSD) (using flash or other low-level technologies) emerge as viable competitors of hard disk drives (HDD). Capacity and access times of storage devices create a trade-off between SSD and HDD. Slow random accesses in HDD have been replaced by efficient random accesses in SSD, but their available capacity is one or more orders of magnitude more expensive than the one of HDD. Indexing, however, is designed assuming HDD as secondary storage, thus minimizing random accesses at the expense of capacity. Indexing data using SSD as secondary storage requires treating capacity as a scarce resource. To this end, we introduce approximate tree indexing, which employs probabilistic data structures (Bloom filters) to trade accuracy for size and produce smaller, yet powerful, tree indexes, which we name Bloom filter trees (BF-Trees). BF-Trees exploit pre-existing data ordering or partitioning to offer competitive search performance. We demonstrate, both by an analytical study and by experimental results, that by using workload knowledge and reducing indexing accuracy up to some extent, we can save substantially on capacity when indexing on ordered or partitioned attributes. In particular, in experiments with a synthetic workload, approximate indexing offers 2.22x-48x smaller index footprint with competitive response times, and in experiments with TPCH and a monitoring real-life dataset from an energy company, it offers 1.6x-4x smaller index footprint with competitive search times as well.
引用
收藏
页码:1881 / 1892
页数:12
相关论文
共 45 条
  • [31] OPTIMAL SEMIJOINS FOR DISTRIBUTED DATABASE-SYSTEMS
    MULLIN, JK
    [J]. IEEE TRANSACTIONS ON SOFTWARE ENGINEERING, 1990, 16 (05) : 558 - 560
  • [32] Na G. - J., 2011, J INFECTION, P111
  • [33] O'Neil P., 1997, SIGMOD Record, V26, P38, DOI 10.1145/253262.253268
  • [34] O'Neil P., 1989, P INT WORKSH HIGH PE, V359, P40, DOI DOI 10.1007/3-540-51085-0_42
  • [35] THE SB-TREE - AN INDEX-SEQUENTIAL STRUCTURE FOR HIGH-PERFORMANCE SEQUENTIAL ACCESS
    ONEIL, PE
    [J]. ACTA INFORMATICA, 1992, 29 (03) : 241 - 265
  • [36] INTERPOLATION SEARCH - LOG LOGN SEARCH
    PERL, Y
    ITAI, A
    AVNI, H
    [J]. COMMUNICATIONS OF THE ACM, 1978, 21 (07) : 550 - 553
  • [37] RAMAKRISHNAN R, 2002, DATABASE MANAGEMENT
  • [38] B+-tree Index Optimization by Exploiting Internal Parallelism of Flash-based Solid State Drives
    Roh, Hongchan
    Park, Sanghyun
    Kim, Sungho
    Shin, Mincheol
    Lee, Sang-Won
    [J]. PROCEEDINGS OF THE VLDB ENDOWMENT, 2011, 5 (04): : 286 - 297
  • [39] The Deletable Bloom Filter: A New Member of the Bloom Family
    Rothenberg, Christian Esteve
    Macapuna, Carlos A. B.
    Verdi, Fabio L.
    Magalhaes, Mauricio F.
    [J]. IEEE COMMUNICATIONS LETTERS, 2010, 14 (06) : 557 - 559
  • [40] Sidirourgos L., 2013, P 2013 ACM SIGMOD IN, P893, DOI DOI 10.1145/2463676