A market-oriented hierarchical scheduling strategy in cloud workflow systems.

A cloud workflow system is a type of platform service which facilitates the automation of distributed applications based on the novel cloud infrastructure. Compared with grid environment, data transfer is a big overhead for cloud workflows due to the market-oriented business model in the cloud environments. In this paper, a Revised Discrete Particle Swarm Optimization (RDPSO) is proposed to schedule applications among cloud services that takes both data transmission cost and computation cost into account. Experiment is conducted with a set of workflow applications by varying their data communication costs and computation costs according to a cloud price model. Comparison is made on makespan and cost optimization ratio and the cost savings with RDPSO, the standard PSO and BRS (Best Resource Selection) algorithm. Experimental results show that the proposed RDPSO algorithm can achieve much more cost savings and better performance on makespan and cost optimization.

SUMMARYIn scientific workflow systems, it is critical to ensure the timely completion of scientific workflows. Therefore, temporal constraints as a type of QoS (Quality of Service) specification are usually required to be managed in scientific workflow systems. Specifically, temporal constraint management includes two basic tasks: setting temporal constraints at workflow build-time and updating temporal constraints at workflow run-time. For constraint setting, the current work mainly adopts user-specified temporal constraints without considering the system performance. Hence, it may result in frequent temporal violations which deteriorate the overall workflow execution effectiveness. As regards constraint updating, although not well investigated, so far is in fact of great importance to workflow management tasks such as workflow scheduling and exception handling. In this paper, with a systematic analysis of the above issues, we propose a probabilistic strategy for temporal constraint management which utilizes a novel probability-based temporal consistency model. Specifically for constraint setting, a negotiation process between the client and the service provider is designed to support the setting of coarse-grained temporal constraints and then automatically derive the fine-grained temporal constraints; for constraint updating, the probability time deficit/redundancy propagation process is proposed to update run-time fine-grained temporal constraints when workflow execution is either ahead of or behind the schedule. The effectiveness of our strategy is demonstrated through a case study on an example scientific workflow process in our scientific workflow system.

No abstract

No abstract