Abstract: Over 446,000 people in the United States fall victim to sudden cardiac arrest annually. A person experiences cardiac arrest when they have no pulse because their heart has suddenly stopped beating. As minutes pass, heart muscle dies and the bodys vital organs are deprived of oxygen rich blood. An AED gives an electrical shock to restart the heart within 2 to 3 minutes and gives the person a 90 percent greater chance of surviving than someone who does not receive the shock. In order to increase the survival chances of a person within the Virginia Tech campus and to decrease the annual victim total, an Android mobile application to help people find the nearest AED on campus was developed.
Poster | Presentation | Final project report | Back to topAbstract: Quality of service (QoS) through a hypervisor is cur- rently an unsolved and unexplored problem. It in- volves keeping track of all disk IO through the hyper- visor while keeping the smallest footprint so as not to hinder the performance of the virtual machines (VMs) and, when necessary, throttling disk access to a designated limit. In this paper, we describe our method of QoS inside the Xen hypervisor. We achieved this goal by modifying the Dom0's backend block driver and providing a new module in the ker- nel of the Dom0. We then proceeded to use account- ing data to throttle the VMs' disk IO to provide a system where basic QoS could occur.
Poster | Presentation | Final project report | Back to topAbstract: The application we built is an AED finder for Virginia Tech campus. Our client, VT Rescue, is responsible for all the AEDs installed throughout the campus buildings and our application will help manage AEDs and make them more accessible by everyone at Virginia Tech. Prior to this project, there were not any known mechanism or an equipment to help people on campus to locate an AED quickly in an emergency situation. Therefore, we believe that this application can help many people survive from a dangerous cardiac arrest situation.
Poster | Presentation | Final project report | Back to topAbstract: We introduce a new software project to test highly available websites and attempt to point out weak points in website infrastructures. This paper will cover the goals of our project, as well as the April 21st, 2011 event that sparked the need for such a thing. The software results of the project can be useful for testing how well a website faces massive failure in the cloud. We also discuss recommendations for developing a highly available in the cloud, with ideas not just specific to the cloud but could apply to a regular data center as well.
Poster | Presentation | Final project report | Back to top