Virtualizing Software and Humans for Elastic Processes in Multiple Clouds-a Service Management Perspective

   page       BibTeX_logo.png   
Schahram Dustdar, Hong-Linh Truong
International Journal of Next-Generation Computing 3(2)

There is a growing trend of combining human-based computation with machine-based computation to solve com- plex problems which cannot be answered with machine-based computation alone. From the computing perspective, integrating machine-based computing elements with human-based computing elements and provisioning them un- der the same model will facilitate the resource elasticity required by complex applications. Although certain works investigate techniques for integrating human-based computing elements with machine-based computing elements, existing computing models for such integrated computing systems are very limited. In fact, the architectures, interconnections, non-functional properties of human-based computing elements are very different from that of contemporary machine-based counterparts. Human-based computing elements are built based on social and bio concepts, thus their architectures, interconnects and non-functional properties are extremely complex and dy- namic, compared with that of machine-based computing elements. In this paper, we examine fundamental issues in virtualizing human-based computing elements and machine-based computing elements using service-oriented computing concepts in order to create highly scalable computing systems of hybrid services to support the elastic- ity of software and people in complex applications. We will outline our Vienna Elastic Computing Model which aims at introducing techniques and frameworks to support multi-dimensional elastic processes atop multiple cloud systems of software-based and human-based services. This paper will analyze several service management issues to support the virtualization of machine-based and human-based computing elements to support such elastic processes.