Simulation and Prototyping of a Coherent Distributed Dynamic Load Balancing Algorithm

This paper presents the performance evaluation of a coherent distributed dynamic load balancing algorithm called the DGP (Decentralised Global Plans) algorithm. The performance evaluation is underpinned by a research methodology integrating a simulation-based performance evaluation with a prototype-based performance evaluation. The DGP algorithm addresses the issue of coherence and coordination in distributed dynamic load balancing algorithms. The DGP algorithm is based on a strategy called Global Plans (GP). This new approach rst explicitly speci es a global goal and then design a strategy around this global goal such that each scheduler (1) takes into account local decisions made by other schedulers; (2) takes into account the e ect of its local decisions on the overall system and (3) ensures load balancing. A simulation-based performance evaluation of DGP with two other well-known dynamic load balancing algorithms published in the literature shows that DGP performs consistently better. More signi cantly, the results indicate that the global plan approach provides a better framework for the design of distributed dynamic load balancing algorithms. A DGP-based prototype has been developed using a distributed programming model which supports dynamic con guration. The overall prototype has been implemented and tested on a network of Sun workstations. The results from the prototype-based performance evaluation are consistent over a range of experiments and clearly demonstrate the bene ts gained from load balancing. An evaluation of both simulation and prototyping results indicated that both set of results were consistent providing a high level of con dence in the results and their applicability.