Implementation of the fault tolerance in computational grid using agents by meta-modelling approach

摘要

In the grid environment, some of the nodes may be working while others may not be active. Alternatively, all the computers could be operational, but their interconnection network may fail. From the perspective of one computer, such network partitioning may appear as a failure to other computers. These types of failures may lead to a major impact on the whole application, which is executing on the Grid for many days. In this paper we will be meta-modelling the computational grid and implementing the fault tolerant mechanism using Java agents. The purpose of the work proposed in this paper is to automate the development of a computational grid and creating graphical workflows of applications using domain-specific modelling techniques. This paper is to provide a high level view for the construction of Grid applications with the flexibility in design and deployment.