Load Balancing Algorithm Improvement Of FastDFS And Application On Website For Soil-water Conservation

Human life and production rely on natural environment on the earth. However, during the transformation of nature, also brought increasing damage to the natural environment. The water and soil erosion is one of the major environmental problems to our nation. The automatic acquisition devices of soil and water conservation monitoring stations upload data document collection to soil and water conservation website, not only can provide historical and current monitoring data, but also can provide data to professional for scientific analysis and study in the area and make corresponding countermeasures for soil and water conservation. So, it is indispensable to construct a complete and efficient distributed file system for the monitoring of soil and water conservation.With the development of information age, more and more data appear with picture and video forms, the traditional storage architecture can't meet our current requirement, we have to use distributed file system. After analyzing the requirement of soil and water conservation website and comparing different kinds of distributed file systems at home and abroad, we decide to use FastDFS as file system to save, access and manage data, and deploy the simulate experimental system on the soil and water conservation website.After studying the current load balancing algorithm of the FastDFS, we propose an improved algorithm for FastDFS. The current algorithm only consider the total storage space of the server group, the improved algorithm consider the performance, the existing storage space and the historical linking number of the server group. The simulation experiments show that the load balancing effect and the I/O performance of experimental system performs better after being improved.The study in this paper comes from the actual requirement of soil and water conservation website. After deploying the improved algorithm for FastDFS on soil and water conservation simulate experimental website, the experiments show that the experimental system performs better after being improved.