Paper Title: Enabling an Optimal Application Ecosystem on ZNS Storage

Paper Abstract: Zoned Namespace Storage (ZNS) leverages a unified host software stack to enable key benefits like reduced over-provisioning and write-amplification along with QoS variability, as compared with conventional flash storage. Applications on ZNS need to be configured optimally to harness these benefits and deliver a sustained transaction throughput and latency for users at scale. An example is web-scale multitenant applications using database management systems at the backend, which utilize storage engines and filesystems configured on ZNS. While trying to enable an application ecosystem on ZNS, users need to deal with multiple components ranging from zoned NVMe devices, associated firmware, tools and utilities, Linux host software working with storage engines, databases, associated plug-ins, and application workloads. This is followed by optimizing the application workloads for a desirable performance and scalability on ZNS, as compared with conventional storage. An easy-to-use ZNS Enablement Kit with scripts and instructions to build, install, configure, optimize, and benchmark an end-to-end application ecosystem on ZNS can accelerate its deployment and deliver an out-of-the-box experience for the user community. The kit is available from WDC for customers who are interested in evaluating ZNS technology. This presentation will describe the content of the kit in detail, and how it can be accessed. As industry collaborations are pursued to enhance and standardize the application ecosystem around Zoned Storage, WDC will publish the kit to a wider community.

Paper Author: Sanhita Sarkar, Global Director, Western Digital

Author Bio: Sanhita Sarkar is a Global Director, Software Development at Western Digital, where she oversees design, development, and benchmarking of storage solutions spanning edge, data center and cloud. She focuses on key vertical markets involving databases, artificial intelligence, streaming and web-scale applications, which encompass Industrial Internet of Things (IIoT), Defense and Intelligence, Media and Entertainment, Financial Services, Genomics, and Healthcare. Sanhita previously held engineering leadership roles at Teradata, SGI, Oracle, and a few startups. She was responsible for delivering optimized software and solutions involving large memory, scale-up, and scale-out systems. Sanhita has authored multiple patents, published several papers, and has been a chair and organizer at several conferences and meetups. She received her Ph.D. in Electrical Engineering and Computer Science from the University of Minnesota, Minneapolis.