Performance evaluation of 3-hierarchical resource management model with grid service architecture

Abstract

The paper constructs a 3-hierarchical resource management model (3-HRMM) in order to adapt to variability of grid resources. We propose resource brokering strategy and adaptive performance task allocation algorithm on 3-HRMM. Resource brokering strategy improves performance through brokering of the application and LRM(Local Resource Manager), a component of 3-HRMM, by performance index and resolves a fault caused by virtual organization by using LRM virtual observer. Under performance evaluation through real-time resource monitoring, adaptive performance task allocation algorithm supports load balancing and fault tolerance. This paper proves feasibility of grid service through the implementation of 3-hierarchical resource management model(3-HRMM), and it shows some results of performance evaluation based on its by using two applications such as image rendering and fractal image processing. © Springer-Verlag Berlin Heidelberg 2007.