The Research of Grid Resource Scheduling Mechanism Based on Pareto Optimality

摘要

As a new computing pattern, Grid computing makes the aggregation, selection, and sharing of geographically distributed heterogeneous resources possible to solve large-scale issues of commerce, engineering, and science. However, it is problematic that the geographic distributed resources owned by different institutions with their different price models, usage policies and changing load. The resource producers and resource consumers have different objects, strategies, and requirement. Meanwhile, the availability of resources and the load on them dynamically varies with time. Hence, resource management in these Grid computing environments is a complicated task. An economic-based method is presented to allocate Grid resource, which is based on Pareto optimality theory and realizes the optimal allocation of Grid resource. And this method can largely avoid a waste of resources. This paper describes a Grid bank model that depends on market mechanism to understand deeply Pareto optimality.