The Community for Technology Leaders
Computer and Information Technology, International Conference on (2005)
Shanghai, China
Sept. 21, 2005 to Sept. 23, 2005
ISBN: 0-7695-2432-X
pp: 339-345
Peter Sloot , Informatics Institute, University of Amsterdam
Hakan Yakali , Informatics Institute, University of Amsterdam
Bob Hertzberger , Informatics Institute, University of Amsterdam
Zhiming Zhao , Informatics Institute, University of Amsterdam
Adam Belloum , Informatics Institute, University of Amsterdam
ABSTRACT
<p>In a Problem Solving Environment (PSE), a scientific workflow management system (SWMS) provides a meta environment for managing activities and data in scientific experiments, for prototyping experimental computing systems and for orchestrating the runtime system behaviour. A Grid infrastructure makes data and computing intensive experiments feasible in PSEs but also requires the management of workflow to support dynamics of the flow execution. A dynamic SWMS includes a human user in the runtime loop of a flow execution, and allows an engine to flexibly orchestrate a workflow according to the human decision and the runtime states of the environment. In this paper, we present our research in an ongoing project: Virtual Laboratory for e-Science (VL-e). An agent based solution is proposed to enhance an existing Grid enabled Problem Solving Environment framework called VLAM-G. The intelligence for problem solving strategies and for workflow orchestration is encapsulated as knowledge in two types of agents: study managers and scenario conductors.</p>
INDEX TERMS
Grid, Problem Solving Environment, Scientific Work.ow, Multi Agent Systems, Petri net.
CITATION
Peter Sloot, Hakan Yakali, Bob Hertzberger, Zhiming Zhao, Adam Belloum, "Dynamic Work.ow in a Grid Enabled Problem Solving Environment", Computer and Information Technology, International Conference on, vol. 00, no. , pp. 339-345, 2005, doi:10.1109/CIT.2005.101
112 ms
(Ver )