基于纠删码技术的云存储数据块部署方案

Erasure Code Based Data Block Deployment Method in Cloud Storage

针对云存储系统中因忽视集群中存储节点之间的差异而引起的存储代价过高、可靠性较低、节点负载能力不足等问题,提出了段排序交换算法(FSSA).首先对数据块部署问题进行数学建模;然后根据各个节点的负载情况进行分段,并在各个分段中依据数据可靠性的需求对节点进行初步选择;最后根据数学模型中对目标函数的分析在分段选择的结果中选取适当的节点进行数据部署.仿真结果表明,采用FSSA算法可以在保证数据存储可靠性的基础之上,降低数据存储代价、增强系统负载均衡能力.

Aiming at the problems of high storage cost,low reliability,and insufficient node load balancing capacity caused by ignoring the difference between storage nodes in a cloud storage,we propose a fragment sort swap algorithm(FSSA).First,we establish a mathematical model for the data block placement problem.Then,we divide the fragments according to the load condition of each node and select the nodes according to the requirements of data reliability in each segment.Finally,according to the analysis of the objective function in the mathematical model,we select the appropriate nodes in the segmentation result to place data.The simulation results show that the proposed FSSA algorithm can reduce data storage cost and enhance system load balancing ability in addition to ensuring the reliability of data storage.