Abstract: |
The goal of the thesis is to explore the major issues in the design and implementation of resource brokers of a Grid system that uses a computational economy driven architecture for managing resources and scheduling task farming applications on large-scale distributed resources. Layered component-oriented modular architectures for resource broker design and implementation will be explored. Deadline and budget constrained scheduling algorithms will be developed and implemented. The efficiency of these algorithms will be estimated experimentally in order to verify their ability in allocating resources depending on their availability, capability, user-level performance, and cost. |