Run-time resolution of service property conflicts in web service composition

Abstract

With rapid development of Web service technologies, service composition has become a common approach to realizing complex business processes. Due to the large number of services developed and deployed independently by various providers, undesirable interactions between properties of different service components may occur when they are composed into a composite service. Such service property conflicts become a serious obstacle for service composition to meet users' requirements. Although traditional feature interaction techniques may prevent some property conflicts in service design stage, many conflicts occur during execution based on certain run-time data; therefore must be resolved online. In this paper, we propose a scheme to address the problem of run-time resolution of service property conflicts. We first formulate the conflict resolution problem with a bi-objective optimization model based on user's revenue, which represents the QoS and success rate of a service. Then we solve the bi-objective optimization model to obtain a set of Pareto solutions and rank the solutions to identify the optimal one, which gives the best roll back strategy and alternative service plan for resolving a service property conflict. We also implement the proposed scheme in a prototype of online shopping application and evaluate performance of the scheme through experiments. The obtained experimental results indicate that our scheme is effective and efficient in resolving service property conflicts at runtime.

An efficient algorithm and tool for detecting dangerous website vulnerabilities

“…For web applications, buffer overflow occurs when a user or a hacker transfers input data beyond the processor's storage capacity, causing the system to crash or execute malicious codes that the hacker inserts (Foster et al, 2005;Xu et al, 2016). There are two types of BoF, including stack-based and heap-based (Kaur and Kaur, 2016).…”