CS Colloquium - Oana Balmau

Event time: 
Wednesday, February 5, 2020 - 4:00pm
AKW 200 See map
51 Prospect Street
New Haven, CT 06511
Event description: 

CS Colloquium - Oana Balmau

Host: Avi Silberschatz

Coffee available at 3:45

Title: Redesigning Storage Systems for Future Workloads, Hardware, and Performance Requirements


The storage stack needs to undergo radical change to keep up with future workloads, hardware, and performance requirements. First, workloads evolved from following a read-heavy pattern (e.g., a static web-page) to a write-heavy profile where the read:write ratio is closer to 1:1 (e.g., users’ actions are being recorded, while they are browsing web-pages). Second, the hardware is undergoing rapid changes.  The divide between fine-grained volatile memory and slow block-level storage is rapidly being bridged by the emerging byte-addressable non-volatile memory devices and the fast block-addressable NVMe SSDs (e.g., Intel Optane NVMe SSDs). Third, performance requirements in storage systems now emphasize low tail latency, in addition to high throughput. In this talk I discuss the main storage systems challenges raised by future workloads, hardware, and performance requirements, and present a new design paradigm for key-value stores to efficiently handle write-heavy workloads. Existing storage systems have fundamental limitations that do not allow them to fully leverage modern hardware. These limitations stem from the longstanding idea that storage systems need to optimize storage device usage. However, new storage hardware shifts the bottleneck from I/O to CPU. Based on this key observation, I introduce a new key-value store design which minimizes the use of CPU, to achieve good performance on block-addressable NVMe SSDs. The overarching principles that optimize CPU use are: (1) avoid maintaining the data sorted on-disk, (2) minimize contention due to shared data structures, and (3) remove expensive maintenance operations ran internally by the key-value store.


Oana Balmau is a PhD candidate in the School of Computer Science at the University of Sydney, advised by Prof. Willy Zwaenepoel. She earned her Bachelors and Masters degrees in Computer Science from EPFL, Switzerland. Her research interests are computer systems and storage technologies. Currently, she is working towards redesigning storage systems, aimed at future workloads and hardware. (Homepage: oanabalmau.com.).