Enterprise Computing and IT Infrastructure Management
Information Technology (IT) infrastructures - computational, storage, and networking resources - in large scale enterprises are constituting a significant portion of their operating budgets. Even though the cost of hardware continues to drop, the bulk of IT budgets are going towards deploying and maintaining these large scale complex infrastructures. The need for easier manageability and lower operating costs has introduced a whole range of exciting computer systems problems that our faculty is currently addressing: self-deploying and configuring software for the environment and applications at hand, self-tuning systems for the best performance at any time, self-healing systems to anticipate failures ahead of their occurrence and reduce down-times, remote monitoring and diagnosis software for enhancing system health and security, power management and temperature aware data center design.
Computer Systems Lab (CSL) - Lead Faculty: Professors Sivasubramaniam and Urgaonkar
The research conducted in CSL lies at the intersection of cutting edge technologies in hardware, networks, systems software and applications towards enhancing performance, manageability, integrity, availability, and security with minimal operational costs. Some of the important contributions to date include scheduling and resource management for large scale clusters, power management of resource-constrained as well as high-end computer systems, techniques for exploiting deep storage hierarchies, provisioning end-to-end Quality-of-Service assurance in computation and storage services, virtualization technologies for easing management of large data centers, and proactive fault tolerance for ensuring higher system availability. Close collaboration with industrial partners have ensured the adoption of ideas in product offerings. Further, many of our evaluation tools, especially in the area of power modeling, are being used by groups at other institutions for their research.
