AGARM: An adaptive grid application and resource monitor framework

Abstract

Most people assume that grid computing is the mainstream high performance and throughput computing platform in the future. To monitor these grid environment and provide grid information service become extremely important for using grid computing resource efficiently. In this paper we propose an adaptive grid application and resource monitor framework named AGARM. The AGARM framework can be used not only to monitor resources but also to monitor applications in the grid environment. Moreover, we show how the monitor activity is optimized by the adaptive pulling techniques, which is based on the moving average estimation algorithm. © Springer-Verlag Berlin Heidelberg 2005.