Run-time Revenue Maximization for Composite Web Services with Response Time Commitments

Abstract

We investigate dynamic decision mechanisms for composite web services maximizing the expected revenue for the providers of composite services. A composite web service is represented by a (sequential) workflow, and for each task within this workflow, a number of service alternatives may be available. These alternatives offer the same functionality at different price-quality levels. After executing a sub-service, it is decided which alternative of the next sub-service in the workflow is invoked. The decisions optimizing expected revenue are based on observed response times, costs and responsetime characteristics of the alternatives as well as end-toend response-time objectives and corresponding rewards and penalties. We propose an approach, based on dynamic programming, to determine the optimal, dynamic selection policy.Extensive numerical examples show significant potential gain in expected revenues using the dynamic approach compared to other, non-dynamic approaches.

Traffic Management for Cloud Federation

“…A number of solutions have been proposed for the problem of dynamic, runtime QoS-aware service selection and composition within SOA [46][47][48][49]. These (proactive) solutions aim to adapt the service composition dynamically at runtime.…”

Analysis of Revenue Improvements with Runtime Adaptation of Service Composition Based on Conditional Request Retries

“…Due to the dynamic service composition it may happen that every composite service request is served by different composition. The service selection is driven by the solution of a suitable optimization problem, which is reduced to the linear optimization problem [9], or the optimization is based on evolutionary computation [10] or is based on the principles of dynamic programming [12]. However, none of [9,10,12] consider the possible applicability of retry mechanisms, i.e.…”

“…How to aggregate some of frequently used workflow patterns and transform the workflow into the sequential one is illustrated, e.g. in [12]. -The selection of candiate services for each task i, i = 1, .…”

“…the optimal compostion does not change at runtime. More recent work in this area focuses on dynamic, runtime composition solutions and adaptations [9,10,12]. For each task invocation, the orchestrator dynamically binds the task of the abstract composition to an actual implementation (i.e., concrete service), selecting it from the pool of service providers that offer it.…”

“…For examples how to calculate the endto-end response time distribution of some other frequently used workflow design patterns, see [12].…”