Visit ShanghaiTech University | 中文
HOME > News and Events > Seminars
The Freeze-Frame File System
Date: 2015/11/27             Browse: 340

Speaker: Weijia Song

Time: Nov 27, 4:00pm - 5:00pm.

Location: Room 306, Teaching Center

Abstract:

Growing numbers of applications require real-time analysis of streaming data. We argue that existing solutions are poorly matched to the need, and introduce our new Freeze-Frame File System. FFFS is able to accept streams of updates while simultaneously materializing read-only past states at arbitrary granularity on demand. FFFS is fast and accurate: we keep the update history in a memory-mapped log, cache recently retrieved data for repeat reads, and use a hybrid of a real-time and a logical clock to respond to read requests in a manner that is both temporally precise and causally consistent. Experiments show that FFFS snapshots achieve 10’s of milliseconds precision. Write throughput can reach 450MBps with a 10G network.

Bio:

Weijia Song is a post doctoral researcher in the Department of Computer Science at  Cornell University. He received his BS in 2001 and MS in 2004 from Beijing Institute of Technology. Then he joined the Computing Center at Peking University as a software engineer. He received his Ph.D. from Peking University in 2014. Weijia is interested in building and optimizing large scale distributed systems.

He seeks new technologies to make cloud systems run smarter. His recent research topics include distributed cloud file system, overlay cloud, and cloud resource scheduling. He has published 14 academic papers in journals and conferences. He holds 5 China patents on distributed and network systems.

SIST-Seminar 15051