Grid computing has emerged as a new paradigm for next generation parallel and distributed computing. The real and specific problem underlying the Grid concept is the availability and sharing of resources in dynamic, multi-institutional virtual organizations. A Grid based system requires heterogeneous resource management and wide-area scalability. Many organizations, based on network-distributed computing systems, require a tlexible overall system configuration that enables site autonomy, multiple scheduling policies and fault tolerance. In this paper, extensive results of simulation experiments, evaluating the scalability degree of batch and online mode scheduling techniques in a data-intensive grid environment are presented. The scenario was studied in view the forthcoming scientific data-intensive applications foreseen by the EUropean DataGrid Project.
|Titolo:||Evaluating Scheduling Algorithms in a Data GRID Network|
|Data di pubblicazione:||2002|
|Nome del convegno:||10th IEEE ICON 2002|
|Appare nelle tipologie:||4.1 Contributo in Atti di convegno|