Course Objectives and Description

An important artifact of the digital millennium is the need to store and operate on vast amounts of data. Technology trends are increasingly allowing users to easily create large amounts of data (e.g., scientific, multimedia, sensor). Legal requirements such as Sarbanes-Oxley act require reliable archival for extended durations. The organization, accessibility, and management of this data is challenging. This advanced graduate course focuses on all aspects of distributed storage; access models, organization and management, security, availability, scalability, reliability, and storage networking. Storage systems covered will include Bayou, BlueFS, DAFS, Elephant, Farsite, Ficus, GoogleFS, Grapevine, HP AutoRAID, IBP, Ivy, LOCKSS, NASD, Oceanstore, PAST, Pangaea, Petal/Frangipani, Samsara, Segank and Zebra. The course will be organized around discussing research papers with significant grades from a course research project.

Prerequisites

Introductory graduate level knowledge of Operating Systems, Distributed Systems and/or Computer Networks. This is an advanced graduate course; students are expected to be highly motivated to read and explore new distributed storage technologies.