An Approach to Data Flow Optimization for False Invoice Emergency Management on Cloud

As the cloud computing has become popular, there is a tendency to develop the service-oriented workflow on cloud. On false invoice emergency management, service-oriented workflow on cloud is a good choice. However, during the workflow execution, large amounts of intermediate data need be stored effectively, otherwise causing unnecessary data transfers, so that become a bottleneck to performance. In this paper, a data scheduling strategy is proposed, which offers additional efficiencies by allowing services of workflow to execution with less data transfers by scheduling the data among data centers based on the data collaboration relations. Our data scheduling strategy not only makes the data transfer time reduced obviously, but also makes the running cost of data scheduling reduced obviously. Simulations demonstrate that the new architecture brings about at least 65% and 47% improvements than existing architecture in process execution time and throughput.