Using partial evaluation in distributed query evaluation

A basic idea in parallel query processing is that one is prepared to do more computation than strictly necessary at individual sites in order to reduce the elapsed time, the network traffic, or both in the evaluation of the query. We develop this idea for the evaluation of boolean XPath queries over a tree that is fragmented, both horizontally and vertically over a number of sites. The key idea is to send the whole query to each site which partially evaluates, in parallel, the query and sends the results as compact boolean functions to a coordinator which combines these to obtain the result. This approach has several advantages. First, each site is visited only once, even if several fragments of the tree are stored at that site. Second, no prior constraints on how the tree is decomposed are needed, nor is any structural information about the tree required, such as a DTD. Third, there is a satisfactory bound on the total computation performed on all sites and on the total network traffic. We also develop a simple incremental maintenance algorithm that requires communication only with the sites at which changes have taken place; moreover the network traffic depends neither on the data nor on the update. These results, we believe, illustrate the usefulness and potential of partial evaluation in distributed systems as well as centralized xml stores for evaluating XPath queries and beyond.