Adaptive query processing and the grid: opportunities and challenges

Abstract

Grid technologies have been developed in response to an increase in demand for computing applications designed to yield the benefits from collaboration, data sharing and sophisticated interaction of autonomous and geographically dispersed resources. Distributed query processing (DQP) is an appealing solution for expressing and efficiently evaluating requests across grid resources. In this paper: (i) we identify parts of the grid infrastructure that facilitate, and open new directions for, query processing over grid-enabled heterogeneous and autonomous databases, stressing the need for adaptive query processing (AQP); (ii) we discuss some basic challenges arising from the new opportunities and outline the unsuitability for use in a grid setting and narrow specialisation of existing proposals for AQP; and (iii) we suggest a generic adaptivity framework as a promising way forward.