摘要
在大数据查询过程中,有时需要对两个有关联的文件同时进行操作,在这种情况下进行缓存时,可以通过考虑文件的关联性以提高查询效率和缓存命中率。提出一种基于文件关联性的协同缓存策略称为CoCache。利用文件的元信息,提出定义文件关联性的方法,并且给出缓存方案,即在进行缓存和替换的时候,将有关联性的部分绑定操作。实验表明,CoCache可以提高数据缓存效率,减少查询时间。
In the process of big data query,it is sometimes necessary to operate two related files at the same time.Therefore,when caching in this case,the query efficiency and cache hit rate can be improved by considering the relevance of the files. This paper proposed a co-caching strategy based on file association called CoCache. Using the meta-information of the file,a method of defining the association of the file was proposed,and a caching scheme was provided,that is,when the caching and the replacement were performed,the associated partial binding operation was performed. Experiments showed that CoCache improved data cache efficiency and reduce query time.
作者
孟子潇
Meng Zixiao(Shanghai Key Lab of Intelligent Information Processing, School of Computer Science, Fudan University, Shanghai 200433, China)
出处
《计算机应用与软件》
北大核心
2018年第7期52-56,共5页
Computer Applications and Software
关键词
大数据
文件关联
协同缓存
Big data
File relevance
Collaborative cache