NERSCPowering Scientific Discovery for 50 Years

Ravi Cheema

Ravi.jpg
Ravi Cheema
Storage Systems Group
National Energy Research Scientific Computing Center
Phone: (510) 495-8016
Mobile: (925) 807-9166
Lawrence Berkeley National Laboratory
1 Cyclotron Road
Mail Stop 59R4010A
Berkeley, CA 94720

Biographical Sketch

Ravi currently works in NERSC Storage Group and shares responsibility to support, deploy and develop NERSC Global Filesystems. He is also the primary point of contact for supporting storage for Joint Genome Institute and NERSC Data Transfer nodes.

Conference Papers

Glenn K. Lockwood, Kirill Lozinskiy, Lisa Gerhardt, Ravi Cheema, Damian Hazen, Nicholas J. Wright, "Designing an All-Flash Lustre File System for the 2020 NERSC Perlmutter System", Proceedings of the 2019 Cray User Group, Montreal, January 1, 2019,

New experimental and AI-driven workloads are moving into the realm of extreme-scale HPC systems at the same time that high-performance flash is becoming cost-effective to deploy at scale. This confluence poses a number of new technical and economic challenges and opportunities in designing the next generation of HPC storage and I/O subsystems to achieve the right balance of bandwidth, latency, endurance, and cost. In this paper, we present the quantitative approach to requirements definition that resulted in the 30 PB all-flash Lustre file system that will be deployed with NERSC's upcoming Perlmutter system in 2020. By integrating analysis of current workloads and projections of future performance and throughput, we were able to constrain many critical design space parameters and quantitatively demonstrate that Perlmutter will not only deliver optimal performance, but effectively balance cost with capacity, endurance, and many modern features of Lustre.

Book Chapters

Glenn K. Lockwood, Kirill Lozinskiy, Lisa Gerhardt, Ravi Cheema, Damian Hazen, Nicholas J. Wright, "A Quantitative Approach to Architecting All-Flash Lustre File Systems", ISC High Performance 2019: High Performance Computing, edited by Michele Weiland, Guido Juckeland, Sadaf Alam, Heike Jagode, (Springer International Publishing: 2019) Pages: 183--197 doi: 10.1007/978-3-030-34356-9_16

New experimental and AI-driven workloads are moving into the realm of extreme-scale HPC systems at the same time that high-performance flash is becoming cost-effective to deploy at scale. This confluence poses a number of new technical and economic challenges and opportunities in designing the next generation of HPC storage and I/O subsystems to achieve the right balance of bandwidth, latency, endurance, and cost. In this work, we present quantitative models that use workload data from existing, disk-based file systems to project the architectural requirements of all-flash Lustre file systems. Using data from NERSC’s Cori I/O subsystem, we then demonstrate the minimum required capacity for data, capacity for metadata and data-on-MDT, and SSD endurance for a future all-flash Lustre file system.

Presentation/Talks

Greg Butler, Ravi Cheema, Damian Hazen, Kristy Kallback-Rose, Rei Lee, Glenn Lockwood, NERSC Community File System, March 4, 2020,

Presentation at Storage Technology Showcase providing an update on NERSC's Storage 2020 Strategy & Progress and newly deployed Community File System, including data migration process.

Ravi Cheema, Kristy Kallback-Rose, Storage 2020 Strategy & Progress - NERSC Site Update at HPCXXL User Group Meeting, September 24, 2019,

NERSC site update including Systems Overview, Storage 2020 Strategy & Progress, GPFS-HPSS-Integration Testing, Tape Library Update and Futures.

Kirill Lozinskiy, Glenn K. Lockwood, Lisa Gerhardt, Ravi Cheema, Damian Hazen, Nicholas J. Wright, A Quantitative Approach to Architecting All‐Flash Lustre File Systems, Lustre User Group (LUG) 2019, May 15, 2019,

Kirill Lozinskiy, Lisa Gerhardt, Annette Greiner, Ravi Cheema, Damian Hazen, Kristy Kallback-Rose, Rei Lee, User-Friendly Data Management for Scientific Computing Users, Cray User Group (CUG) 2019, May 9, 2019,

Wrangling data at a scientific computing center can be a major challenge for users, particularly when quotas may impact their ability to utilize resources. In such an environment, a task as simple as listing space usage for one's files can take hours. The National Energy Research Scientific Computing Center (NERSC) has roughly 50 PBs of shared storage utilizing more than 4.6B inodes, and a 146 PB high-performance tape archive, all accessible from two supercomputers. As data volumes increase exponentially, managing data is becoming a larger burden on scientists. To ease the pain, we have designed and built a “Data Dashboard”. Here, in a web-enabled visual application, our 7,000 users can easily review their usage against quotas, discover patterns, and identify candidate files for archiving or deletion. We describe this system, the framework supporting it, and the challenges for such a framework moving into the exascale age.

J. Hick, R. Lee, R. Cheema, K. Fagnan, GPFS for Life Sciences at NERSC, GPFS User Group Meeting, May 20, 2015,

A report showing both high and low-level changes made to our life sciences workloads to support them on GPFS file systems.

Reports

GK Lockwood, D Hazen, Q Koziol, RS Canon, K Antypas, J Balewski, N Balthaser, W Bhimji, J Botts, J Broughton, TL Butler, GF Butler, R Cheema, C Daley, T Declerck, L Gerhardt, WE Hurlbert, KA Kallback-Rose, S Leak, J Lee, R Lee, J Liu, K Lozinskiy, D Paul, Prabhat, C Snavely, J Srinivasan, T Stone Gibbins, NJ Wright, "Storage 2020: A Vision for the Future of HPC Storage", October 20, 2017, LBNL LBNL-2001072,

As the DOE Office of Science's mission computing facility, NERSC will follow this roadmap and deploy these new storage technologies to continue delivering storage resources that meet the needs of its broad user community. NERSC's diversity of workflows encompass significant portions of open science workloads as well, and the findings presented in this report are also intended to be a blueprint for how the evolving storage landscape can be best utilized by the greater HPC community. Executing the strategy presented here will ensure that emerging I/O technologies will be both applicable to and effective in enabling scientific discovery through extreme-scale simulation and data analysis in the coming decade.