A Cross-Layer Approach to Performance Monitoring of Web Services

An increasing amount of applications are currently built as Web Service compositions based on the TCP/IP+HTTP protocol stack. In case of any deviations from desired runtime-behavior, problematic Web Services have to be substituted and their execution plans have to be updated accordingly. One challenge is to detect deviations as early as possible allowing timely adaption of execution plans. We advocate a cross-layer approach to detect bad performance and service interruptions much earlier than by waiting for their propagation through the full protocol stack. This position paper describes an approach to gain detailed real-time information about Web Service behavior and performance based on a cross-layer analysis of the TCP/IP+HTTP protocols. In this paper we focus especially on TCP. The results are used to make decisions supporting service selection and replanning in service-oriented computing scenarios. Furthermore, generic architectural components are proposed implementing the functionality needed which can be used in different web-based scenarios.