Queuing Mechanism in Migrating Workflow System Based on Cloud Computing Paradigm (MWfSCC)

Abstract

Migrating workflow system based on cloud computing paradigm can be regarded as a type of platform service which facilitates the automation of distributed large-scale e-business and e-science applications in the cloud computing platform. CMI is a software entity that can migrate from one site to another in order to satisfy its goals in cloud workflow system. However, if a large number of CMIs migrate to one work machine within a short period of time, congestion is so serious that the average executive time for CMIs would be much longer due to the competition for the limited resources. In this paper, we introduce the queuing conception into cloud workflow system to solve this problem. Two queuing models are created for corresponding two different types of work machine, called sole work-node model and compound work-node model. For each model, a set of detailed queuing rules, service rules and the CMI dispatching mechanism are defined. We choose P/N/m/∞ queuing model to implement queuing mechanism in sole work-node model, and use dynamic priority queue to manage waiting CMI. Compound work-node is made up of a lot of sole work-nodes, when a CMI migrated to one compound work-node, it would be dispatched to one sole-work node according to the task type to be executed. The effectiveness of the queuing mechanism is evident in that it can alleviate the load of work nodes, which are illustrated with the experiment results at the end of paper.