Performance and Scalability of XML Query Processing

This paper discusses basic issues about the performance of semi-structured query processing in very large data sets. It is based on recent algorithm-engineering work, on the state of the art in performance management for XML query processing and on theoretical studies about the complexity structure of the querying problem. Its main conclusions provide a concrete view on the interaction between terabyte scale XML data, query complexity and current or future computer architectures. To provide a concrete and synthetic view of this diverse body of knowledge, the presentation follows a fictional use case whose characters face query problems of varying complexity that are set in multiple contexts and analyzed today before being projected to 2020 and 2030.