EECS Department Colloquium Series
Wednesday, January 21, 2015
As more and more data moves to the cloud, data replication has become prohibitively costly and there is an acute need for efficient, fault-tolerant schemes for data storage. Coding theory offers solutions for fault-tolerant storage that are potentially far more efficient than replication. At the same time, the cloud storage setting presents some unique challenges that traditional error-correcting codes do not handle. There have been some novel solution concepts proposed to address these challenges (such as Regenerating Codes and Locally Repairable Codes).
In this talk we will describe the challenges, both theoretical and practical, in designing efficient erasure coding schemes for cloud storage. Our case study will be Locally Repairable Codes, or LRCs, that were first deployed by Azure Storage in 2012, resulting in tremendous savings in hardware costs, and have since been deployed in other Microsoft products. These codes are inspired by Locally Testable and Locally Decodable codes from theoretical computer science, and provide efficient recovery (independent of the code length) for typical failure scenarios.
This talk is based on joint work with collaborators from MSR and the Azure storage team. No prior background will be assumed.
Parikshit Gopalan in a researcher at Microsoft in the Windows Azure storage team. Prior to this, he was a researcher at MSR Silicon Valley. He received a Ph.D from Georgia Tech and a B. Tech from IIT Bombay, and has held postdoctoral positions at UT Austin and the University of Washington. He is interested in theoretical computer science, focusing on Computational Complexity and Coding Theory. On the applied side, he is interested in fault-tolerant storage using erasure coding. His work in this area has been deployed in several Microsoft products and has won awards including the 2014 IEEE Communication and Information Theory Society Joint Paper Prize, the 2013 Microsoft Technical Community Network Storage Technical Award and the USENIX ATC 2012 best paper award.
|Return to EECS Joint Colloquium|