Seminar: Peripherals No More
Sam H. Noh
Professor, Dept. of Computer Science & Engineering
Dean, Graduate School of Artificial Intelligence
UNIST (Ulsan National Institute of Science & Technology)
Monday, February 14, 2022
1100 Torgersen Hall
We are in the midst of a dramatic change in what computer systems look like. Our traditional view of computer systems composed of the CPU, main memory, and very slow storage device has recently been challenged with the advent of SSDs. Today, with the advent of persistent memory (PM), which has characteristics of DRAM as well as storage, we are, seeing an even more dramatic change in what computers look like.
In this talk, I will share experiences and results from work that we, at the NECSST (Next-generation Embedded/Computer System Software Technology) lab at UNIST, have been conducting in regards to these changes. In the first part of the talk, I will describe a recent study that we conducted that revisits an old topic, that is, the buffer cache. Specifically, we consider the use of PM as a buffer cache and consider how the persistent characteristics of PM can be exploited to improve storage I/O performance. By harnessing this PM layer, we show that legacy file systems and traditional storage devices can be supported, while still performing comparably with state-of-the-art PM based file systems such as NOVA and Ext4-DAX. Furthermore, compared to Ext4, our approach can outperform Ext4 by more than 9 x, while providing durable in-order file system semantics, whereas Ext4 cannot.
Time permitting, I will also revisit RAID, another traditional storage topic. Here, we consider how today's All-Flash-Array (AFA) storage systems should be configured in light of modern technological changes. Recent flash SSDs provide high I/O bandwidth such that the storage bandwidth can easily surpass the network bandwidth by aggregating a few SSDs. However, with traditional RAID, it is challenging to unlock the full performance of SSDs, the main source of performance degradation being garbage collection (GC). We find that existing AFA designs are susceptible to GC at SSD-level and AFA software level. To alleviate the performance interference caused by GC at both levels, we take a spatial separation approach that partitions SSDs into the front-end SSDs dedicated to serve write requests and the back-end SSDs where GC is performed. Our approach allows the storage bandwidth to match the full network performance without being interfered by AFA-level GC. We provide evaluations that show that our approach is effective for a variety of workloads.
Sam H.(Hyuk) Noh received the BE degree in computer engineering from the Seoul National University, Seoul, Korea, in 1986, and the PhD degree from the Department of Computer Science, University of Maryland, College Park, MD, in 1993. He held a visiting faculty position at the George Washington University, Washington, DC, from 1993 to 1994 before joining Hongik University, Seoul, Korea, where he was a Professor in the School of Computer and Information Engineering until the Spring of 2015. Starting from the Fall of 2015 he joined UNIST (Ulsan National Institute of Science and Technology), a young Science and Tech focused national university, where he is a Professor at the Department of Computer Science and Engineering and the inaugural Dean of the Graduate School of Artificial Intelligence in the College of Information and Biotechnology (and also former Dean of the School of Electrical and Computer Engineering). From August 2001 to August 2002, he was also a visiting associate professor with the University of Maryland Institute of Advanced Computer Studies (UMIACS), College Park, MD.
He has served as General Chair, Steering Committee (Chair) Member, Program Chair, and Program Committee Member on a number of technical conferences and workshops including USENIX ATC, USENIX FAST, ACM Eurosys, USENIX HotStorage, ACM EMSOFT, ACM SOSP, IEEE RTAS, ACM ASPLOS, USENIX OSDI, ACM LCTES, IEEE ICPADS, and WWW, among others. He has also been serving as Editor-in-Chief of the ACM Transactions on Storage since the summer of 2016. His current research interests include operating system issues pertaining to computer systems with a focus on the use of new memory technologies such as flash memory and persistent memory. He is a Fellow of the ACM and IEEE and a member of USENIX and KIISE.