MooseFS中chunkserver负载均衡算法研究

Chunkserver load balancing selection algorithm on MooseFS

作为云存储的核心基础平台,分布式文件系统的重要性日益凸显。分布式文件系统中数据存储在多台计算机节点上,必然会出现负载均衡问题。首先,对MooseFS的系统架构进行了研究,然后分析了MooseFS分布式文件系统中chunkserver选择算法,研究了chunkserver算法的负载均衡性能,最后对其进行了改进。经过实验测试对比,实验结果显示改进算法能显著提高chunkserver的负载均衡性能。

As the core foundation platfoml of the cloud storage, distributed file system is more and more important.In the distributed file system, data is stored in multiple computers, the appearance of the load balancing problem is inevitable. Firstly, research the architecture of the MooseFS system, and then research the load balancing performance of ehunkserver algorithm. Finally, modify the chunkserver algorithm. Alter comparison of the experimental tests, the experimental results show that the modified algorithm can significantly inlprove the load-balanced performance of the ehunkserver.