Model evolution by runtime parameter adaptation

Models can help software engineers to reason about design-time decisions before implementing a system. This paper focuses on models that deal with non-functional prop-erties, such as reliability and performance. To build such models, one must rely on numerical estimates of various pa-rameters provided by domain experts or extracted by other similar systems. Unfortunately, estimates are seldom cor-rect. In addition, in dynamic environments, the value of pa-rameters may change over time. We discuss an approach that addresses these issues by keeping models alive at run time and feeding a Bayesian estimator with data collected from the running system, which produces updated param-eters. The updated model provides an increasingly bet-ter representation of the system. By analyzing the updated model at run time, it is possible to detect or predict if a de-sired property is, or will be, violated by the running imple-mentation. Requirement violations may trigger automatic reconfigurations or recovery actions aimed at guarantee-ing the desired goals. We illustrate a working framework supporting our methodology and apply it to an example in which a Web service orchestrated composition is modeled through a Discrete Time Markov Chain. Numerical simula-tions show the effectiveness of the approach. 1