GSAF: A Grid-based Services Transfer Framework

In recent years, traditional computing systems face the problems of scalability as the need for information processing services is ever increasing. Grid, as a pool of computing resources, solves the problem in some degree by providing an integrated computing and resources environment. Thus, there emerge many strategies related to grid resource allocation management (GRAM). However, almost all of these strategies focus on how to allocate the existing computing resources but overlook the possibility of controlling these resources. Therefore, services and computing nodes are always tightly coupled. In this paper, we describe a new framework called GSAF (a grid-based services transfer framework) to solve the problem. GSAF binds services and computing nodes dynamically through services migration and service cache to break the tight coupling. The migration of service components to more efficient computing nodes helps balance the processing burden and extends the computing abilities. In our experiment, we provide a prototype application of GSAF in the data mining system to demonstrate it.