A two level approach for managing resource and data intensive tasks in grids
In recent years, Grid computing has emerged as an attractive platform to tackle various large-scale problems, especially in the field of science and engineering. Scheduling Grid resources involves a number of challenging issues, mainly due to the distributed and dynamic nature of the Grids. This paper focuses on the resource allocation for a particular type of resource intensive tasks called Processable Bulk Data Transfer (PBDT) tasks in a Grid environment. The defining trait of a PBDT task is a large raw data-file at a source node that needs to be processed in some way before it can be used at a set of sink nodes. Our scheduling approach uses a Bi-level decision-making architecture. This paper analyzes the performance of the proposed architecture at various workload conditions. This architecture can be extended for other types of tasks using the concepts presented.