面向网格计算的动态容错服务框架设计（英文）

Design on framework of dynamic fault tolerance service faced to grid computing

网格计算在研究复杂问题求解和解决大型科学计算方面有重要应用,针对高度异构网格本身导致的容错服务难题,探讨了一种面向网格计算的动态容错服务排序框架设计。分析和总结了网格计算的容错要求,给出了基于网格环境的相关故障定义,建立了故障的分类模型,提出了一种包括电网故障检测和故障管理内容在内的动态容错服务框架,最后给出了详细的故障容错服务流程。借助仿真实验,初步验证了所提出设计框架的合理性和可行性。研究结果表明:该设计可为面向网格计算的动态容错服务提供一个新的参考框架。

Aimed at the puzzle of fault tolerance service resulted in highly heterogeneous grid itself, this paper ex- plored a sort of design on framework of dynamic fault tolerance service faced to grid computing. In the paper, the fault tolerance requirements of grid computing was analyzed and summarized, and the definition of the relevant fault was presented based on the grid environment. The classification model of the fault was established in this pa- per and a kind of dynamic fault tolerance service framework was put forward in which the grid fault detection and fault management was included. The function of each component was described one by one in the grid framework, and finally, the detailed flow of fault tolerance service was presented. The elementary simulation experiment par- tially demonstrated the feasibility and rationality of the proposed framework. The research results show that it could provide a new reference framework for dynamic fault tolerance service faced to grid computing.