Small file storage optimization model in cloud environment
Xiaoling Dong, Xiaoyu Yin, Dongbo Yu, Min Zhang · 2022 IEEE 10th Joint International Information Technology and Artificial Intelligence Conference (ITAIC) · 2022
Aiming at the problems of metadata server performance bottleneck and low file reading efficiency when the storage system in the cloud environment faces small file storage. Starting from the inherent data association between small files, this paper extracts the association through a lightweight pattern matching algorithm. The small files are merged on the basis of this feature, which improves the rationality of the merged files, and saves the small files in the same merged file into the client cache when the file is read to improve the cache read hit rate. Experiments verify that the scheme in this paper can effectively improve the access efficiency of small files.