Recently, grid-computing has been paid attention to the public, but many people still thinks that grid-computing is to combine several personal computers into a supercomputer.
Grid-computing is not only to use available CPUs(Central Processing Unit), but also to share separate data and storage resources in addition to human resources and knowledge.
Korean Ministry of National Defense has designated GIG(Global Information Grid) based on integral model for implementing NCW(Network-Centric Warfare), which means the intelligence war in modern war trend.
NCW(Network-Centric Warfare) make a difference in the military efficiency by sharing the intelligences from the enhanced communication capability of high performance computer processing units. Ministry of National Defense in both Korean and USA has covered wider range in simulated war based on the grid-computing. However, there are still challenges due to the difference middle ware structure in grid-computing environment and a uniform scheduling.
Therefore, it is important to have a working scheduling method to improve the resources efficiencies by recovering and reallocation of the resources under error conditions.
In this thesis, firstly the research background for grid-computing and their classifications are provided. Then, the grid structures in both Korea and USA are analyzed, and the adopting-dynamic reallocation scheduling method are presented for a new grid computing infrastructure.
It is expected that the proposed method make the resources efficiencies higher by resource allocation and recovering under error conditions.