Data Lab is currently located in Meshel Hall 359. Data symbolizes Data-Oriented Science. The research in this lab not only includes big data but also includes databases, data analysis, cloud computing, etc. Date-oriented science studies data of any size, any format, anytime, and any source.
Our computers are cloud clusters built on private hosted cloud clusters and running open-source cloud platforms. In these clouds, we created powerful virtual servers to support our research. By the power of cloud computing model, these servers are built by on-demand project requests and destroyed after the projects are finished. We can even build virtual clusters within the clouds and easily achieve scalability and high-availability.
Our lab currently has two cloud clusters running. The first cloud (Sarah Cloud, named after my daughter) is made of 4 high-end Dell tower servers, including 76 Xeon E5 vCPU cores, 448 GB high-speed RAM, 29TB RAID-5 distributed storage, and a 5TB NFS shared drive. It also includes a NVidia K4000 GPU for computation intensive experiments. This cloud creates many virtual servers for our students in the lab to develop their research projects.
The second cloud is an experimental cloud (Atom Cloud) made of normal PCs. Currently, we have 10 Dell Precision T3500 and 4 Dell Precision T3400 workstations. The purpose of this cloud is to let students test different cloud computing models and build clouds from bare-metal machines. We have previously succeeded to put both virtualization services and big data services on the same cloud cluster sharing slave machines with different master nodes.
Our door opens to every student interested in data-oriented science. We have many students who have successfully conducted their graduate projects and theses in this lab. Research work produced in this lab has been introduced in the special topic course on cloud computing (CSCI5859P) since 2014.