Time-to-market and cost control demand driving complexities out of today's mission-critical, large-scale enterprise software applications. In the meanwhile, SLAs require adequate performance built in upfront to meet the stringent business operational requirements. These forces have led to a more disciplined approach to building software by adopting Service-Oriented Architecture (SOA). In this paper, we present an analytic model for predicting the performance of SOA-based enterprise software applications. We demonstrate how emerging SOA frameworks such as XML Web services can be characterized with well-known performance laws.
展开▼
