论文部分内容阅读
传统分布式文件系统中的副本量控制策略,主要是从内部资源的角度考虑副本的创建,未考虑外部需求的变化。这种策略不适合部署在以服务为基础、内部存储资源丰富的智慧城市云存储中心上。提出了一种结合数据安全(最小副本量)和服务需求(最优副本量)的副本量控制模型。模型中采用基于双时间粒度预测算法预测文件流行度,依据文件的流行度和系统资源动态调整云中副本数量。仿真实验结果表明,本文采用的动态副本机制在应对突发的大规模并发访问方面有显著优势;同时,相比常规的基于访问频率的动态副本,本文策略的存储资源占用量更少。
The traditional distributed file system copy volume control strategy, mainly from the perspective of internal resources to consider the creation of a copy, did not take into account changes in external demand. This strategy is not suitable for deployment in the service-based, internal storage resource-rich smart city cloud storage center. A copy volume control model combining data security (minimum copy size) and service requirement (optimal copy size) is proposed. The model predicts file popularity based on dual-time granularity prediction algorithm and dynamically adjusts the number of copies in the cloud according to the popularity of the file and system resources. The simulation results show that the dynamic replication mechanism adopted in this paper has significant advantages in coping with sudden large-scale concurrent accesses. Meanwhile, compared with the conventional dynamic replicas based on access frequency, the proposed strategy consumes less storage resources.