Dynamic economic- and QoS-based approach for workflow-oriented distributed services allocation

With the increasing scale of distributed services in wide-area environment, there are hundreds of thousands of accessible services, which are deployed in heterogeneous, geographically distributed service systems. The processing of services in wide-area environment is also becoming more and more intricate, which promotes the need for linking large-scale distributed computational resources and heterogeneous data resources. The service providers have different price policies and cost models for their services, and varying loads and availability, and also considering about the QoS requirements of service clients, it would require support for dynamic allocation during the process of long running services to achieve the economic cost. In this paper, we propose a dynamic economic- and QoS-based approach for distributed services allocation, which is workflow-oriented. We take the service request as workflow process, which would be achieved by series of element services. The allocation of such element services would be decided dynamically due to the economic objective function (EOF) to address the issues of flexibility, availability, performance, economic cost and QoS requirements in the execution process of services.