A Head Record Cache Structure to Improve the Operations on Big Files in Cloud Storage Servers

Cache and prefetching is now widely used in storage systems for disks to speed up accessing data. Although cache and prefetching are well accepted technologies in storage field, they are unsuitable for cloud storage systems because most of requests in cloud is for large files, which leads to undesirable hit rate and speed performance. To solve this issue, an improved head record cache (HRC) structure model is proposed in this paper based on reshuffling disk cache structure and prefetching technologies, aiming at improving reading performance in a cloud environment. Compared to previous researches, this model has better read performance in a cloud environment, since HRC increases hit rate. The experimental results demonstrate that the system has 18% better reading performance than traditional cloud storage system.