An extended multi-layered blackboard approach for the optimization of world-spanning query execution plans

Abstract

New application domains ranging from the area of computational science (e.g. CERN Datagrid) to international business processes (e.g. Wal-Mart logistics) demand ever increasing storage capacities on a global scale. Therefore the ability to query and access data efficiently is one of the major challenges that drive research and development in the area of wide distributed data management. Due to the situation of a huge variety of underlying heterogeneous and homogenous infrastructures, where each showing specific hard- and software resources and a number of possible data sources, the optimization of the access process to the desired data is a challenging task. Hence, this paper introduces a novel heuristic approach for the optimization of query execution plans (QEP) on a global scale based on an extended multi-layered blackboard mechanism to determine which available (1) data, (2) resources, and (3) operations to use. The proposed method is illustrated by an application scenario of finding an optimal QEP for the retrieval of music information from an online music store.