Research On Grid Resource Management And Allocation Based On Benefit Optimization

With the rapid development of internet technology, the dispersed computers to realize scientific computing alone become very difficult. In order to meet user the computing resources and the growing demand for computing power, grid computing model comes into being. Grid is an integrated resource sharing and collaboration environment. In order to achieve the sharing of computing resources, storage resources, data resources, knowledge resources and expert resources, the grid connects dispersed computers, storage equipment, and scientific equipment together with network, and integrates into huge virtural super computer. Grid resource management model is an important part of the grid. It realized for distributed, heterogeneous, dynamic grid resources management. Grid resource scheduling algorithm plays an important role in achieving the grid resource reasonable scheduling and satisfying customer's demand.This thesis analyses the present research situation of the grid, and summarizes the deficiencies of the grid resource management model. Trying to find a more reasonable grid resource management model from the angle of grid development. The main work and innovation are listed as follows:1.This thesis introduces the grid research status and the key technology. It analyzes the grid system structure and resource scheduling algorithm, and summarizes the deficiencies of the grid resource management model.2.According to the grid resource management model's characteristics and deficiencies. From the angle of the customer satisfaction, this thesis presents an benefit optimization model for grid resource management. This model improves the users and the grid resource providers'participation in the grid.3.Analyzing the deficiencies of the current grid scheduling algorithms, this thesis presents a grid scheduling algorithm should be implemented in phases. In the early in order to attract more resource's providers to join in the grid, auction-based benefit optimization scheduling algorithm should be used. In the balanced development phase of the grid, in order to improve the grid customer's satisfaction, the scheduling algorithm based on benefit function drive should be used.4.The time constraint factorαand cost constraint factorβare introduced into the grid scheduling algorithm. Users can define the values ofαandβthrough their preferences which reflect the grid users'different preferences for the time and cost.5.This thesis introduces the features and the characteristics of the simulation toolbox gridsim, and simulates the scheduling algorithm using the toolbox gridsim. Experimental results show that the presented algorithm is feasible, which can improve the grid resource management and customer satisfaction.