重复数据删除系统元数据存储布局研究

发布时间：2018-03-05 22:28

本文选题：重复数据删除　切入点：元数据　出处：《北京理工大学学报》2014年05期 　论文类型：期刊论文

【摘要】：根据重复数据删除的特点,针对连续数据访问为主的应用,设计基于B+树的元数据索引结构;将元数据分为冷热两种,分别设计不同的元数据布局策略.研究结果表明,低能耗磁盘能够保持更长时间的待机状态以便节能,同时可以有效地减少重复数据删除时的随机磁盘访问次数.并且没有额外的硬件开销,实现仅在磁盘上存储索引,而不是之前研究中的在磁盘和内存中存储索引.
[Abstract]:According to the characteristics of repeated data deletion, the index structure of metadata based on B-tree is designed for the application of continuous data access, and the metadata is divided into two types: hot and cold, and different metadata layout strategies are designed. Low-power disks can maintain longer standby state to save energy, and can effectively reduce the number of random disk visits when data is deleted repeatedly. There is no additional hardware overhead, so the index can only be stored on disk. Rather than storing indexes in disk and memory as previously studied.
【作者单位】：北京理工大学计算机学院;北京物资学院信息学院;
【基金】：国家自然科学基金资助项目(61003262) 北京市“青年英才计划”资助项目(YETP1532)
【分类号】：TP333

【参考文献】