科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2014年
10期
73-75
,共3页
云存储%布隆过滤%漂移数据%删除算法
雲存儲%佈隆過濾%漂移數據%刪除算法
운존저%포륭과려%표이수거%산제산법
cloud storage%bloom filter%drift data delete algorithm
云系统中的漂移数据作为大量冗余数据的一种,对其有效及时的删除能保证云存储系统的稳定与运行。传统的云系统漂移数据删除技术采用全文件静态分块策略,由于操作粒度较小,对漂移数据的删除率不高。提出一种基于垂直分层布隆过滤的云系统漂移数据删除算法,设计基于客户端/服务器的云存储系统漂移数据删除机制体系构架,采用布隆过滤技术进行数据源端处理,引入热度垂直索引热度来衡量数据块边界偏移所造成用户的访问集中热点损失,文件根据内容划分成可变长度的数据块,进行垂直分层,得到备份集中数据对象的粒度,采用奇异值分解的方式对漂移数据序列的细节信号展示,根据矩阵奇异值分解矢量的唯一性,避免一些数据块边界偏移造成的误删和漏删。仿真实验表明采用该算法进行云系统的漂移数据删除,性能较好,执行效率和精度优越于传统算法。
雲繫統中的漂移數據作為大量冗餘數據的一種,對其有效及時的刪除能保證雲存儲繫統的穩定與運行。傳統的雲繫統漂移數據刪除技術採用全文件靜態分塊策略,由于操作粒度較小,對漂移數據的刪除率不高。提齣一種基于垂直分層佈隆過濾的雲繫統漂移數據刪除算法,設計基于客戶耑/服務器的雲存儲繫統漂移數據刪除機製體繫構架,採用佈隆過濾技術進行數據源耑處理,引入熱度垂直索引熱度來衡量數據塊邊界偏移所造成用戶的訪問集中熱點損失,文件根據內容劃分成可變長度的數據塊,進行垂直分層,得到備份集中數據對象的粒度,採用奇異值分解的方式對漂移數據序列的細節信號展示,根據矩陣奇異值分解矢量的唯一性,避免一些數據塊邊界偏移造成的誤刪和漏刪。倣真實驗錶明採用該算法進行雲繫統的漂移數據刪除,性能較好,執行效率和精度優越于傳統算法。
운계통중적표이수거작위대량용여수거적일충,대기유효급시적산제능보증운존저계통적은정여운행。전통적운계통표이수거산제기술채용전문건정태분괴책략,유우조작립도교소,대표이수거적산제솔불고。제출일충기우수직분층포륭과려적운계통표이수거산제산법,설계기우객호단/복무기적운존저계통표이수거산제궤제체계구가,채용포륭과려기술진행수거원단처리,인입열도수직색인열도래형량수거괴변계편이소조성용호적방문집중열점손실,문건근거내용화분성가변장도적수거괴,진행수직분층,득도비빈집중수거대상적립도,채용기이치분해적방식대표이수거서렬적세절신호전시,근거구진기이치분해시량적유일성,피면일사수거괴변계편이조성적오산화루산。방진실험표명채용해산법진행운계통적표이수거산제,성능교호,집행효솔화정도우월우전통산법。
The cloud drift data in the system has a large amount of redundant data, so the effective and timely remove can guarantee the stability and operation of cloud storage system. The cloud drift data of traditional delete technology uses full file static block strategy, due to the operation of smaller particle size, so the drift data deletion rate is not high. A removing algorithm of cloud drift data is proposed based on vertical layered bloom filter design in cloud storage system, the client/server drift data delete system based on the introduction of vertical heat is introduced, the heat index is used to measure the block boundary offset caused by user access concentrated hot loss, the documents according to the content is divided into data blocks of variable length, size of the backup set of data objects is computed, and the vertical stratification is obtained, the singular value decomposition method is used for the drift data sequence of the detail signal display, according to the ma-trix singular value vector of the uniqueness of decomposition, it can avoid some block boundary offset remove and drain de-lete. Simulation results show that the new drift data removing method has good performance, the efficiency and accuracy of the algorithm is superior to the traditional method.