Publication Date
7-2006
Abstract
Several scientific projects use BOINC (Berkeley Open Infrastructure for Network Computing) to perform large-scale simulations using volunteers� computers (workers) across the Internet. In general, the scheduling of tasks in BOINC uses a First-Come-First-Serve policy and no attention is paid to workers� past performance, such as whether they have tended to perform tasks promptly and correctly. In this paper we use SimBA, a discrete-event simulator of BOINC applications, to study new threshold-based scheduling strategies for BOINC projects that use availability and reliability metrics to classify workers and distribute tasks according to this classification. We show that if availability and reliability thresholds are selected properly, then the workers� throughput of valid results increases significantly in BOINC projects.
Comments
UTEP-CS-06-30