A distributed and cooperative load balancing mechanism for large-scale P2P systems

This paper proposes a distributed and cooperative scheduling mechanism for dynamic load-balancing on a large-scale distributed computing environment. In the proposed mechanism, the scheduling processes are performed by independent distributed schedulers on individual computing resources. Decentralized mechanisms are more suitable for dynamic load-balancing of a large-scale distributed computing environment than centralized mechanisms in terms of scalability and fault tolerance. Experimental results show that the proposed scheduling mechanism has high scalability and efficiency, without any excessive concentration of processing even if the number of computing resources increases