Complex distributed systems are critical in the modern computing experience, yet many undergraduate computing curricula give students little exposure to the full complexity of these systems. Investigators have developed a prototype gaming environment, called DWORLD, to provide students more realistic experiences with distributed systems. The current project extends the capabilities of this environment. Students compete by addressing a variety of issues in a virtualized distributed system in order to keep the system healthy and operating normally. Game scenarios are designed to enhance student skills and knowledge of key concepts such as scalability, robustness, fault-tolerance and the ability to build a distributed system. Expected outcomes include a software package that can be easily adopted by other faculty and a collection of lab modules and assignments.
The game is being piloted in undergraduate courses at Polytechnic University of New York and CUNY Brooklyn College. Workshops are used to train faculty how and why to use the gaming environment in their courses. As part of a broader context, the project assesses the potential for games as an important tool in STEM education.
Complex distributed systems are a fundamental building block of computing in the 21st century. As more computing moves to mobile devices and the cloud, the systems will continue to increase in importance. Therefore it is an important national need to develop educational tools that will enable universities and companies to educate the next generation of technology professionals to build and evolve these distributed systems. The goal of this project was to evaluate scientifically the efficacy of a simulation game about distributed systems (DWORLD). In this game students, working in teams, are cast in the role of administrators trying to solve real-time problems in a complex distributed system. The intellectual merit of the project was grounded in its novel approach of using techniques from game design to enhance student engagement and learning, and the broader impact was anticipated to be a tool that would further student learning in this area with associated empirical evidence that it did so. During the project an examination was developed that characterized student learning in distributed systems. A prototype version of DWORLD was developed prior to this grant project. During this project significant development was done on DWORLD, to make it more robust and to integrate troubleshooting scenarios into the game that matched specific learning goals. The survey was used in two versions of an undergraduate course at Polytechnic University, and the robustified version of the game was deployed in one offering of that course and one offering of a course at Brooklyn College. Approximately 60 students gained experience using DWORLD. Unfortunately, due to personal transitions mid-grant period of several members of the project team, the project was not completed.
