With the rapid development of cloud environment, the capabilities of systems have been promoted with powerful computing and storage. But for the characteristic of “pay-as-you-go” of cloud resources, it is necessary ...With the rapid development of cloud environment, the capabilities of systems have been promoted with powerful computing and storage. But for the characteristic of “pay-as-you-go” of cloud resources, it is necessary to consider the different data storage cost. Especially for processing of “old data” in longterm storage, an appropriate strategy is needed to reduce users’ cost. Considering the characteristics of price stratification in the current commercial cloud environment, a three-level price stratified storage strategy is proposed based on the CTT-SP algorithm, which stores part of the “old data” on relatively inexpensive secondary and tertiary storage, and ensures that the time delay caused by three-level storage does not exceed the deadline. Compared with other storage methods, the experimental result shows the strategy proposed can guarantee the time delay while reducing the cost of users significantly in longterm storage.展开更多
基金Anhui Natural Science Foundation 1908085MF206 and National Natural Science Foundation of China (NO. 61402007, 61573022)the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry.
文摘With the rapid development of cloud environment, the capabilities of systems have been promoted with powerful computing and storage. But for the characteristic of “pay-as-you-go” of cloud resources, it is necessary to consider the different data storage cost. Especially for processing of “old data” in longterm storage, an appropriate strategy is needed to reduce users’ cost. Considering the characteristics of price stratification in the current commercial cloud environment, a three-level price stratified storage strategy is proposed based on the CTT-SP algorithm, which stores part of the “old data” on relatively inexpensive secondary and tertiary storage, and ensures that the time delay caused by three-level storage does not exceed the deadline. Compared with other storage methods, the experimental result shows the strategy proposed can guarantee the time delay while reducing the cost of users significantly in longterm storage.
