The operation of modern distributed enterprises, be they commercial, scientific, or health related, generate massive quantities of data. Decision makers increasingly utilize On- Line Analytical Processing (OLAP) tools to glean from this rich data resource nuggets of information which can be used to better run their enterprises. A typical approach to OLAP is to construct a single centralized data repository by copying all of the raw data from the sites where it is generated to a cental location, where it is integrated, and then to route all queries to that central location. As the amount of data and number of sites and users grows this approach suffers from significant scalability problems.

In this paper, we present a model and algorithmic framework for an "OLAP-Enabled Grid" whose goal is the efficient support of OLAP operations. We show how a Grid computing infrastructure can be used to store and manage expensive to compute data aggregations and to answer OLAP queries in a fully distributed manner. Our focus is on the efficient optimization of resources for answering queries based on a distributed query algorithm which uses cached and pre-aggregated data stored over a Grid computing infrastructure.