An Agent-based Resource Allocation Model for computational grids

• Published in: Multiagent and grid systems 2005-01, Vol.1 (1), p.17-27
• Main Authors: Manvi, S.S., Birje, M.N., Prasad, Bhanu
• Format: Article
• Language: English
• ISSN: 1574-1702; 1875-9076
• DOI: 10.3233/MGS-2005-1103

The paper presents an Agent-based Resource Allocation Model (ARAM) for grid computing. Three types of agents, namely Job Agents (JAs), Resource Brokering Agents (RBAs), and Resource Monitoring Agents (RMAs), are used. The JAs are mobile and they represent the users and perform tasks such as completing the job requests of the users and executing the jobs at suitable grid nodes. The RBAs are static and they act as resource schedulers as well as brokers for the users to submit their jobs through JAs. The RBAs incorporate an economic model and a queuing model. The RMAs are static and they reside in the nodes of the local clusters and inform the status of the resources to the local cluster servers. The model differs from other existing models, as it has the following characteristics: 1) topology-based migration of the agents, 2) analysis of different types of migrations according to the topology and the agents' overheads, and 3) resource allocation based on dynamic pricing and negotiation. The model is evaluated in a simulated environment to investigate the behavior of several parameters.