Course: EEL6871 Autonomic Computing
Abstract
This course introduces beginning graduate
students to key concepts and techniques underlying the design and
engineering of autonomic computing and networking (AC) systems. AC
systems are IT systems capable of self-management, self-healing,
self-tuning, self-configuration and self-protection. Course content
includes an introduction to the defining characteristics of AC, why it is
necessary, foundational AC principles based on control theory, artificial
intelligence and systems concepts. Also covered are case studies and
technologies used to implement AC systems. Recent papers will be
used to discuss integrated systems and methods for AC.
Intellectual Merit
This project will cover system modeling and controller design techniques in autonomic computing. Also students will be exposed to Hadoop framework, big data analysis and have chances to fine tune Hadoop performance through a closed-loop feedback system.
Broader Impact
The results and observations can be used by all scientific researchers that are interested in a self-managing Hadoop system.
Use of FutureGrid
We will use FutureGrid as the testbed for the course project. Students will apply futuregrid Account through the portal and create their own hadoop clusters. On these clusters, they will run several data analysis applications and do some close-loop performance tuning tasks.
Scale Of Use
We will have 12 students each of whom will have a hadoop cluster with approximately 5-10 VMs. Students will perform some Hadoop performance tuning tasks on these clusters.