Detecting Race Conditions in Web Services

Feature interaction has been first formally identified as a problem in the telecommunication domain. With the rapid growth and application of Web Services, feature interaction problem also arises in the dynamic Web Service interactions. Race conditions, one type of functional feature interactions in Web Services, will lead to the unexpected system behavior in the composite Web Services system. This paper proposed a Petri net-based method to detect the race conditions in Web Services. The method can be divided into two phases: modeling Web Services and detecting the race conditions. On the basis of modeling Web Services, the Petri net analysis methods are employed in the detection algorithm. This static detection method will help avoid race conditions in the composite Web Services.