The Research On Cache Strategy In Multiple-cloud Storage System

With the advent of big data era, public cloud storage is widely used for its low-cost,high-capacity and scalability advantages, but it cannot guarantee the quality of service and data security, enterprises build private cloud storage in order to ensure the safety and reliability of the data, but the fluctuations of users' demand, often insufficient or excess storage capacity. Combining private cloud storage which is safe, reliable, fast, and public cloud storage which has low-cost, high-capacity, scalability, multiple-cloud storage comes into being. On the other hand, cloud storage access is limited by network bandwidth and fluctuations, resulting in a long data access latency, cache can reduce network latency in multiple-cloud storage system and provide users a better storage service experience.For cloud storage access highly latency and high cost, design and implement a multiple-cloud storage system to improve cost effective, which builds a highly scalable multiple-cloud storage management platform that manages multiple public clouds and private clouds, and can concurrent upload, download and delete files. Designing multiple-cloud monitoring and storage scheduling scheme, monitoring each cloud storage services, designing rational scheduling policy based on users' needs and service characteristics of different cloud storage to provide users cost-effective storage services. to save download costs, improve download bandwidth, combining with the local space and private cloud storage space, designing data cache; we present least weight replace cache replacement algorithm(Least Weight Replace, LWR), considering file frequency, the elapsed access time of file, file size, file download price, in order to calculate the weight of the file, the algorithm meets users' characteristic.The experimental results show that the multiple-cloud storage system can achieve the basic functions of upload, download and delete files with high concurrent; at the same time we carry out performance comparison test of the proposed cache strategy, the results show that the cache strategy which uses LWR cache replacement algorithm is better than traditional cloud cache policy as the file hit rate and saving the cost.