LD_B^+树:闪存系统上一种DBMS索引结构  

LD_B^+ Tree: DBMS Index Structure over Flash Memory Storage Systems

作　　者：李国徽[1] 赵培[1] 高胜[1] 

机构地区：[1]华中科技大学计算机科学与技术学院,武汉430074

出　　处：《小型微型计算机系统》2012年第2期193-198,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60873030)资助;国家"八六三"高技术研究发展计划项目(2007AA01Z309)资助;国家"十一五"国防预研基金资助

摘　　要：以往的研究大多针对文件系统,而DBMS存在更多细粒度的更新.本文综合考虑闪存自身的特点、设备种类繁多及不同闪存设备读写特性差别大等,提出了一种基于闪存的DBMS索引结构:LD_B+树.LD_B+树根据工作负载的读写特性动态地调节索引模式使之能够适应于不同种类的闪存设备.LD_B+树采用日志结构组织结点,通过结点转换表和日志缓冲区维护索引结构.模拟实验结果表明,不同闪存设备及工作负载下,LD_B+索引结构比B+树和日志型B+树(BFTL)具有6%-63%的性能提高.Flash memory has been widely used in the storage system. Most researches mainly focuses on the file system, However, DBMS has more fine-granularity updates. Considering a wide range of flash memory devices have far different access characteristics, a new DBMS index structure, referred to as LD_B＋ tree, is proposed in the paper, which redefines the basic indexing operation such as retrieval, insertion, deletion, and so on. LD_B＋ tree mainly uses node translation table and logging buffer to maintain the indexing structure. A set of simulation experiments was conducted and the results showed that the LD_B＋ tree outperformed the existing index structure 6% -63% under different workloads and flash packages.

关 键 词：LD_B+树 日志 结点转换表 Merge操作 在线算法 恢复 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]