Erasure Coding for Data Center Storage
主 講 人：Patrick P. C. Lee The Chinese University of Hong Kong
報(bào)告時(shí)間：2018年7月3日（周二）上午 10：00-12：00
報(bào)告地點(diǎn)：計(jì)算所四層報(bào)告廳
報(bào)告摘要 (Abstract)
Data center storage systems increasingly adopt erasure coding to reduce the storage overhead of traditional 3-way replication. However, maintaining high performance in erasure-coded data center storage systems still has many design trade-off issues. In this talk, I will present new results on how to seamlessly adapt erasure codes into data-center architectures, with emphasis on taking into account the data-center-specific characteristics such as hierarchical topologies, bandwidth diversities, and mixed failure modes. In particular, I will present novel techniques and code constructions that improve repair performance, update performance, and storage efficiency of data center storage. Our methodologies have been validated on Hadoop and Quantcast File System platforms, and we have released various open-source prototypes on our designs.
主講人簡(jiǎn)介 (Short Bio)
Patrick P. C. Lee received the Ph.D. degree in Computer Science from Columbia University in 2008. He is now an Associate Professor of the Department of Computer Science and Engineering at the Chinese University of Hong Kong. He currently heads the Applied Distributed Systems Lab and is working very closely with a group of graduate students on different projects in large-scale software systems. His research interests are in various applied/systems topics on improving the dependability of large-scale software systems, including storage systems, distributed systems and networks, and cloud computing.