Overview and Empirical Research on File Correlation in Cloud Storage

Abstract

Data is stored in the cloud, and applications remain on traditional servers. This is also a common architecture that appears in the current process of migration from traditional application architecture to cloud computing architecture. This makes the efficiency and speed between traditional application servers and cloud storage a performance bottleneck for the entire system architecture. At present, the design of almost all cache systems in cloud storage addresses the performance bottlenecks of the system architecture based on the local characteristics of access, that is, the use of the temporal and spatial local characteristics of cloud storage to design the cache system. This article summarizes the prefetching technology and caching technology in cloud storage cache from the perspective of memory hierarchy, and verifies the temporal and spatial locality characteristics of cloud data access through the access of NASA-HTTP log data set.