Usually terms such as high performance and high availability are addressed by big corporations and institutions; however, something has changed over the past years as a real revolution is emerging from university classrooms. Dr. Jose Luis Vazquez-Poletti, Assistant Professor in Computer Architecture at Complutense University of Madrid, describes some of the promising work being carried out by his students.
Cloud computing represents an incredible boost for research in the HPC and HTC areas. Their adoption has increased thanks to this technology, and accordingly, so has the number of worldwide students with a focus on cloud-enabled technology.
Master thesis projects are not only the final challenge for our students, they are also a great opportunity to shine a spotlight on their efforts. This is crucial for attracting interesting job opportunities once they leave the university. Companies demanding technologically qualified employees see in these projects the perfect companion in their recruiting processes, as the candidates can easily showcase their competence with bleeding edge technologies.
Year after year, my students have found in the cloud a powerful tool to make the most of their master thesis projects. This on-demand resource makes it possible to easily develop incredible frameworks for handling real life problems with HPC and HTC… in just an academic year!
This year I have been honored to advise some interesting projects, which are representative of the above statements. In particular, these projects respond to three critical areas that make their way to the media headlines nowadays: communications security, emergency medical services and P2P digital currencies.
On-demand secure teleconferencing on public cloud infrastructures (2012-2013)
One of the main handicaps of migrating HPC and HTC applications to the cloud is security. Having no control over how our data travels through the Internet and not knowing who can access them, makes users reluctant to adopt a public cloud migration strategy. Communications over the Internet are a concrete example where users may need to communicate in a secure way not susceptible to eavesdropping or interception.
Bernardo Pericacho, UCM Master in Computer Science Research.
This project was carried out by Bernardo Pericacho, who was my student during the Master in Computer Science Research offered by our faculty. He proposed two novel secure teleconferencing architectures deployed in the cloud, Teleconferencing Secure Cloud VPN Architecture and Hidden Teleconferencing Server Cloud Architecture, to address possible security breaches and attacks in Voice over Internet Protocol (VoIP) communications in an unsafe environment such as the Internet.
Diagrams of the architectures proposed by the on-demand secure teleconferencing on public cloud infrastructures project.
In addition, a model represented by a decision tree was proposed, in which a user can easily determine which architecture and infrastructure is the one that best fits his needs when establishing a secure teleconference server on a public cloud.
CloudMiner (2013-2014, ongoing)
Nobody would argue that mining bitcoins or one of the many other digital currencies is a great example of HPC gone monetary. Problems arise when having heterogeneous mining rigs and non-monitored maintenance costs. This challenge was taken up by Tomás Restrepo, Juan Arratia and Arturo Pareja, all of them doing their Master Thesis in Computer Science.
CloudMiner team: Tomás Restrepo, Juan Arratia and Arturo Pareja at Madrid’s Mining Engineering Museum.
CloudMiner is a project that aims for a better exploitation of the existing hardware, achieving economical benefit through the mining of virtual crypto-currencies. The main idea is to create a cloud computing resource pool, composed by diverse machines under potentially different architectures. This pool is to be monitored in real-time, enabling the user to start/stop mining at any given point on any of the available machines, also providing information on their current status. Additional options are also possible, like adding or removing resources (supported architectures). Artificial Intelligence based on statistics can be used in order to allow automated control of the mining cloud. These statistics are also visible to the user, to aid decision-making.
Diagram showing CloudMiner features.
EMS in Cloud (2013-2014, ongoing)
In the world of medical emergencies, every second is valuable. Medical teams have mobility needs, and one way to improve their performance is to take advantage of software applications based on cloud systems. This approach makes protocols and tools available to every team who needs them. This was the starting point of Alejandra González, Ricardo Champa and Claudia Montero, also doing their Master Thesis in Computer Science.
EMS in Cloud team: Alejandra González, Ricardo Champa and Claudia Montero at different hospitals and medical facilities in Madrid.
EMS in Cloud is a Software as a Service (SaaS) system deployed on multiple platforms. The framework provides an economical solution, with a focus on service and continuing care.
With EMS in Cloud, the user can easily define and manage medical protocols to be used during emergency operations. These protocols use different metrics and provide action plans according the available medical supplies. If a given medical product is depleted or its composition changes, the user is then requested to modify the affected protocols, increasing the effectiveness of service.
Even if they can be shared, each user can have his own set of protocols. By using the cloud at a Platform as a Service level, both scalability and availability are guaranteed.
Diagram showing EMS in Cloud features.
During the emergency service, the user accesses the protocols through a mobile client, which is able to work offline with most of its functionality when no Internet is available.
The team is currently working in partnership with Manuel Vazquez, medical staff at the Spanish EMS. He is not only providing his knowledge, he will be also testing EMS in Cloud on the field.
Manuel Vazquez, the expert sharing his knowledge with the EMS in Cloud team.
The Future for the Future
The three projects described here have some aspects in common. First, each one works on an interesting challenge and provides a solution for a given real life problem.
Second, they all use the cloud as the cornerstone of their proposed architectures, allowing an easy adoption of the HPC and HTC paradigms where necessary.
Third and the most important, these projects are carried out by students. In a world dominated by the pessimism due to the financial crisis, students are our best asset. Is in this asset in which we need to invest the most.
Students are aware of the harsh situation they will find once they leave the classroom. They are also aware that only ideas will allow them to survive out there. As you have seen in these examples, they have many ideas and, allow me to say, very good ones! They only need to access the technology that will turn these ideas into reality and fill the ranks of the next generation of “HPC/HTC/Cloud shapers”.
Curious about the rest of the projects? Then take a look at http://dsa-research.org/jlvazquez/students/
About the Author
Dr. Jose Luis Vazquez-Poletti is Assistant Professor in Computer Architecture at Complutense University of Madrid (UCM, Spain), and a Cloud Computing Researcher at the Distributed Systems Architecture Research Group (http://dsa-research.org/). He is (and has been) directly involved in EU funded projects, such as EGEE (Grid Computing) and 4CaaSt (PaaS Cloud), as well as many Spanish national initiatives.