Scale-Down Methods for Optimizing Resource Allocation In Providing Virtual Laboratory Environment by Cloud Computing

摘要

Nowadays, Cloud Computing plays a vital role in providing virtual infrastructure for computing demands and services such as Amazon EC2, Google Cloud, Azure. In addition, Cloud Computing has potential for supporting experimental environment as Virtual Laboratory in terms of education. Instead of the real laboratory, Virtual Laboratory can save deploying time and cost. With the model that we are implementing on our High Performance Computing system, lecturers are considered as users who request the virtual resources for experimental environment. Hence, the problem that we pose is how to optimize the mechanism of allocating resources efficiently for both the user side and the server side. This paper contributes two scale-down algorithms which serve Virtual Laboratory users and optimize the scheduler of resource allocation on server side for multiple courses. The results illustrate that our aforementioned algorithms not only save costs by scaling down virtual resources but also satisfy students experiment in using the Virtual Laboratory.