Research Of Grid Resource Management And Allocation Based On Auction Mechanism

The rapid growth of the technology of computing and network mostly promotes the scientific research and commercial application based on Internet. In many areas, there are more and more higher request for the computing capacity and it is difficult for a single computer to undertake calculation task. As the result of the increasing demand for calculative capacity and resource, Grid technology emergence. The final goal of Grid technology is to realize resource share and coordinating work in virtual network environment and eliminate the resource isolated island. Grid core service is an important part of grid computing. As a centrum of entire grid computing, grid core service connect grid under-stratum groupwares to top groupwares and take charge entire grid system in order to ensure grid system works effectively. It is important to research grid core service technology. Resource scheduling technology is a part of grid core service technology. An advanced resource scheduling policy in favor of reducing total run time and total expense in grid computing, by this means enhance entire grid system's capability. Resource management and one of the key research fields scheduling under grid environments of grid systems.This paper aims at how to allocate the Grid resources effectively and reasonably. The author's main workings are below:1,We discuss the grid concept, introduce two popular grid architecture, we analyze emphatically three models of Grid Resource Management, which are Hierarchical Model, Abstract Owner Model, and Computational Market Economy. The characteristics of them are discussed and we get a conclusion.2, We analyze the current economic model. And combine with the characteristics of grid resource allocation, we propose grid resource management model based on auction mechanism.3, We analyze the factors that affect grid system capability and how they affect each other. Scheduling policy must reference any information about simulation system when policy make decision, so we must test some parameters of simulation system after build it.4> We propose a time optimization algorithm based on computing economy, this algorithm that schedules a user application on auction-based resource allocation systems. These allocation systems, which are based on proportional share, allow users to bid higher in order to gain more resource shares. And a broker adjusts a user bidding price periodically on these systems in order to finish the application within a specified deadline and budget, and achieve optimization of the resources allocation.5> Several grid emulators are compared in this paper, and GridSim------a gridmodeling and emulation toolkit is introduced in details. Finally, with the help of GridSim toolkit, the emulation of grid resource scheduling with is realized, and evaluate the performance and accuracy of these algorithms.