This study explores clustering technologies in Higher Vocational Education and Training (VET), specifically at the Malta College of Arts, Science and Technology (MCAST). The goal is to optimize computing resources, given that many students now prefer their personal devices over the institution’s lab equipment, leading to underutilized resources. Whilst VET institutions invest heavily in the latest technologies, the reality they face is that most of the equipment is seldom used by students and most of the hardware is idle and unused for long periods of time. The institute used in this study encompasses 1,500 students and 200 lecturers, offers IT-related courses across EQF Levels 2 to 7 and houses 45 labs equipped with standalone computers.

A survey revealed that students prefer using personal devices due to flexibility and administrative restrictions on lab PCs. However, many face limitations with personal hardware, citing insufficient resources, compatibility issues, and security concerns. They expressed a need for a system offering 32GB RAM, 1TB SSD, powerful processing capabilities, and remote access. Lecturers noted similar issues with the current lab infrastructure, pointing out slow PCs, network limitations, and the absence of essential software tools. They suggested that a new system should provide modern processors, 32GB RAM, support for Linux and Windows, and compatibility with specialized curriculum software.

Based on this feedback, a test lab was set up using OpenStack and the existing lab hardware, comparing its performance and capabilities against VMware ESXi and the current standalone setup. Virtual machines (VMs) were configured across Ubuntu Desktop, Ubuntu Server, and Windows Server to simulate the institute’s environment. Performance tests, including synthetic (CPU, memory, storage, and network), scalability (vertical and horizontal), and real-world tests, were used to assess the system. Real-world scenarios included educational applications like Jupyter Notebooks, SQL queries, GNS3 projects, and Unity game development, reflecting actual course needs.

Results showed that OpenStack and KVM hypervisor, met the academic requirements by creating a scalable, secure, and cost-effective cloud environment. This solution leverages the institution’s existing hardware to support specialized software and provide reliable, high-performance resources for students. Compared to proprietary options like VMware ESXi, OpenStack offered similar performance with greater flexibility and lower costs. It also allowed remote 24/7 access, addressing student needs for convenient and reliable resources that bridge personal computing limitations.

OpenStack demonstrated strong scalability, handling peak workloads effectively and adapting to fluctuating academic demands. However, implementing and managing this system requires specialized open-source expertise, a skill set that may be challenging and costly to obtain. The study was limited by a small-scale lab setup, restricted remote access due to security policies, and hardware constraints, such as the lack of dedicated GPUs. Essential features like high availability and containerization were not explored.

Overall, this study highlights OpenStack as a viable platform for VET institutions seeking to optimize resource usage and improve student access to powerful computing resources, supporting practical and industry-aligned education.

Keywords: Technology, education, cloud computing, virtualisation.