Loading…
This event has ended. Create your own event → Check it out
This event has ended. Create your own
View analytic
Wednesday, March 11 • 4:20pm - 5:10pm
Reducing File System Tail Latencies with Chopper - Jun He, University of Wisconsin-Madison

Sign up or log in to save this to your schedule and see who's attending!

Jun He (UW-Madison) will present Chopper, a tool that efficiently explores the vast input space of file system policies to find behaviors that lead to high-tail latencies. Jun He will focus specifically on block allocation, as complex allocation policies can lead to surprising and costly performance problems. The approach utilizes sophisticated statistical methodologies to explore the search space efficiently while still ensuring excellent coverage of system responses. Chopper has been applied to study the over all behavior of two file systems, and to study Linux ext4 in depth. Chopper is able to identify four internal design issues in ext4 which form a large tail in its performance distribution. By removing the underlying problems in the code, consistent and satisfactory file layouts are achieved.

Speakers
avatar for Jun He

Jun He

PhD Student, University of Wisconsin-Madison
Jun He is a PhD student at Department of Computer Sciences, University of Wisconsin-Madison. He works with Prof. Andrea C. Arpaci-Dusseau and Prof. Remzi H. Arpaci-Dusseau. Recently, he has become interested in taking advantage of statistics techniques, especially the Design Of Experiments (DOE) and Sensitivity Analysis, to guide computer system designing, which could help developers to avoid unexpected behaviors and make more confident... Read More →


Wednesday March 11, 2015 4:20pm - 5:10pm
Loft 2