云計算中 HDFS 數(shù)據(jù)存儲策略研究

打開文本圖片集

摘要：針對在云計算中HDFS存儲小文件數(shù)據(jù)讀取率低、時間長和內(nèi)存消耗大等缺點，文章提出了一種分布式緩存方案策略，通過每個DataNode節(jié)點緩存處理客戶端頻繁訪問的小文件，從而顯著降低了DataNode磁盤訪問的頻率，提高了數(shù)據(jù)的存儲效率，以期為云計算中的存儲技術(shù)提供科學(xué)參考。

關(guān)鍵詞：云計算；HDFS；小文件；方案策略

中圖法分類號：TP333文獻(xiàn)標(biāo)識碼：A

Research on HDFS data storage strategy in cloud computing

XU Kefei

（Jinan Software Research Institute of China United Network Communication Co.，Ltd.，Jinan 250100，China）

Abstract：In view of the shortcomings of low data reading rate， long time and large memory consumption of small files are stored in HDFS in cloud computing， a distributed caching scheme strategy is proposed. Each DataNode caches and processes small files frequently accessed by clients， which significantly reduces the frequency of DataNode disk access and improves the storage efficiency of data， in order to provide scientific reference for storage technology in cloud computing.

Key words： cloud computing， HDFS，small file， scheme strategy

近年來，隨著計算機和互聯(lián)網(wǎng)應(yīng)用技術(shù)的快速發(fā)展，云計算作為一種新的計算模式應(yīng)用而生。（剩余3528字）