Trust Based Grid Scheduling Algorithm for Commercial Grids

Abstract

Commercial grids are the use of grid computing within the context of a business or enterprise. In Commercial grids, users are regularly engaged with the distributed resource providers with whom they have little or no prior experience. The interaction between these distributed resource providers requires resource management and scheduling solutions. The allocation and scheduling of applications on a set of heterogeneous, dynamically changing resources is a serious problem. Current Grid schedulers schedule a given job according to the availability and the performance provided by resources at that moment without considering the previous knowledge of their performance. We have proposed a trust model and evaluated the trust based on affordability, success rate and bandwidth. This proposed trust model has been incorporated with Gridway meta scheduler and tested. While scheduling, the resources with higher trust values are selected by this proposed model. It has been observed that the success rate of the jobs submitted by the user is very much increased by this proposed trust based scheduling.