Abstract:

Teaching of Computer Science require training environments that resembles the real-life situations that the students will face in their professional activity. This is a common practice in other disciplines, like medicine where courses often include training sessions in real-life scenarios. Unfortunately, more often than not our students do not have access to such training environments; and so their education presents severe limitations in this aspect.

Probably, this situation is caused by the need to provide the students with total control over the computational resources. And this requirement entails several problems from the computer laboratory point of view, namely:

* Security. Total control of a system constitutes a potential risk not only for the computer laboratory machines, but also for the users and hosts of the same network. So students are often left in limited environments.

* Environment Complexity. Usually computer labs are shared by several courses or require non trivial configurations. Although automatic installation tools are available, the hosts have to be often manually tuned by technical staff. So, the computer lab can not be re-installed to clean up between sessions.

As a result, students learn in limited, isolated environments that can not be modified; and so some exercises can only be done from a theoretical point of view. Moreover, the above limitations also result in an inefficient organization of the computer labs, e.g. tight timetables for practicing or limited resources for a given course...

Recently, the dramatic performance improvements in hypervisor technologies have made possible to experiment with virtual machines as basic building blocks for flexible computational platforms. This technology is also being used to improve the management of computer labs, see for example [1-4]. Virtualization has also brought about a new computing model, called Cloud computing, for the on-demand provision of virtualized resources as a service. The Amazon Elastic Compute Cloud is probably the best example of this new paradigm for the elastic capacity providing.

In this work, we propose a system that leverages Cloud computing and virtualization technologies for the on demand provisioning of virtualized computer labs. The system presents the following benefits from an educational point of view:

* Real-life training, without the associate security risks.

* Rapid deployment of complex training environments, so the student can freely interact and break the training system.

* Computer infrastructure optimization. The virtualized lab workstations are provided from a centralized pool, and so lab resources can be low-end computers, the energy consumption can be optimized or the computational resources can be outsourced from public providers, among others.

* A computer lab anywhere and anytime. Instructors can easily arrange a practical session so any classroom can be turned in a computer laboratory, designed to fit the course needs.

References:
[|] "Outsourcing a Computer Lab" Don Hardaway, Computer
[2] "Virtualizing an IT Lab for Higher Education Teaching" Nilsgentschen Felde, Tobias Lindinger
[3] "VLabNet: A Virtual Laboratory Environment for Teaching Networking and Data Communications" Valerie J et al.
[4] "Using Virtual Machines to provide a secure Teaching Lab environment", Harry Bulbrook

Keywords:

cloud computing, virtual laboratories, e-learning tools provisioning.