QoS-Oriented Reputation-Aware Query Scheduling in Data Grids
Authors
Abstract
In the last few years, the Grid technology has emerged asan important tool for many scientific and commercial global organiza-
tions. In grid-based systems, intelligent job scheduling is used to achieve
Service Level Objectives (SLOs) and to provide some kind of Quality
of Service (QoS) differentiation between users or applications. In data
grids, the grid infra-structure is used to provide transparent access to
geographically distributed data, which may be replicated in order to
increase availability and performance. In this work, we deal with QoS-
oriented query scheduling in data grids. Although there exist several
works on job scheduling in Grids, QoS-oriented query scheduling in grid-
based databases is still an open issue. For instance, how can we provide
guarantees against response-time expectations? Our proposal uses a rep-
utation system to answer this problem satisfactorily. We also present
experimental results that prove the benefits of proposed strategies.