Scalable filtering of XML data for Web services

With the advent of XML as the de facto language for data interchange and the rapid emergence of Web services, Web applications have to deal with increasing amounts of XML traffic, such as SOAP requests. In order to scale to large number of clients and high data traffic, back-end systems to which Web services are mapped face the challenging task of efficiently filtering, classifying, and routing incoming XML requests to the appropriate component in the infrastructure. As the destination of each XML request depends on its actual type and content, Web service applications have to incorporate highly-efficient content-based routing technology. In this paper, we address the issue of scalable filtering of XML data. We describe sophisticated techniques for matching data against large number of tree-structured filters expressed using the XPath language. We propose a hierarchical XML routing architecture that supports extremely high loads, and we present experimental results that demonstrate its performance.