Our lab includes two Dell blade server chassis, each with 14 blade servers, one of which is used as a starting point for our research and education needs.  One rack has over 1.5 TB of RAM, and the other around 768 GB of RAM.  Each of these chassis has multi-terabyte SANs with close to 30 TB free.  These racks are using the VMware ESXi hypervisor, and we have experimented on a small scale with VMware's SERENGETI which allows SAN storage to be addressed as local storage, allowing HADOOP to perform the map/reduce process on the blades and SAN storage vice on single host/local storage systems.  These blade systems are expandable with room for additional RAM, additional storage, and 3 more blade servers.

You may recall the recent Capstone Report by Chiefs Doucet, Dontchev, Burden, and Skoff, advised by Professor Gondree and Nguyen; Big Data Analytics Test Bed.  They built a VMware SERENGETI cluster which performed quite well.  Their system was a traditional server cluster rather than a blade, but showed that a hypervisor based cluster would work, and work well.

A new product, VMware Big Data Extensions is showing even more promise and gives performance on par with traditional standalone installations, and supports the familiar tools and services common to distributions by Apache HADOOP, Cloudera, Pivotal, Hortonworks, and MapR.  Pig, Hive, and HBase are also supported allowing a single cluster to support multiple distributions.

In addition to being able to provide the platform and create and support classes and labs, Professor Kamel has interest in certificates, short courses, and ultimately a degree in Data Science.  I have supported him in labs in the past in data base and web courses, and we have discussed such collaboration in Big Data as well.  This makes good use of our facilities and our faculty in IS.