• Big Data
    Interagency Working Group
    (BD IWG)

    The Big Data Interagency Working Group (BD IWG) works to facilitate and further the goals of the White House Big Data R&D Initiative.

  • Cyber Physical Systems Interagency Working Group (CPS IWG)

    The CPS IWG is to coordinate programs, budgets, and policy recommendations for Cyber Physical Systems (CPS) research and development (R&D).

  • Cyber Security and Information Assurance Interagency Working Group (CSIA IWG)

    Cyber Security and Information Assurance (CSIA) Interagency Working Group coordinates the activities of the CSIA Program Component Area.

  • Health IT R&D
    Interagency Working Group

    The Health Information Technology Research and Development Interagency Working Group coordinates programs, budgets and policy recommendations for Health IT R&D.

  • Human Computer Interaction & Information Management Interagency Working Group (HCI&IM IWG)

    HCI&IM focuses on information interaction, integration, and management research to develop and measure the performance of new technologies.

  • High Confidence Software & Systems Interagency Working Group (HCSS IWG)

    HCSS R&D supports development of scientific foundations and enabling software and hardware technologies for the engineering, verification and validation, assurance, and certification of complex, networked, distributed computing systems and cyber-physical systems (CPS).

  • High End Computing Interagency Working Group (HEC IWG)

    The HEC IWG coordinates the activities of the High End Computing (HEC) Infrastructure and Applications (I&A) and HEC Research and Development (R&D) Program Component Areas (PCAs).

  • Large Scale Networking Interagency Working Group
    (LSN IWG)

    LSN members coordinate Federal agency networking R&D in leading-edge networking technologies, services, and enhanced performance.

  • Software Productivity, Sustainability, and Quality Interagency Working Group (SPSQ IWG)

    The purpose of the SPSQ IWG is to coordinate the R&D efforts across agencies that transform the frontiers of software science and engineering and to identify R&D areas in need of development that span the science and the technology of software creation and sustainment.

  • Video and Image Analytics
    Interagency Working Group (VIA IWG)

    Formed to ensure and maximize successful coordination and collaboration across the Federal government in the important and growing area of video and image analytics

    VIA CG
  • Wireless Spectrum Research and Development Interagency Working Group (WSRD IWG)

    The Wireless Spectrum R&D (WSRD) Interagency Working Group (IWG) has been formed to coordinate spectrum-related research and development activities across the Federal government.


File:Suggested Research and Development Topics for 2005-2009.pdf

Jump to: navigation, search
Suggested_Research_and_Development_Topics_for_2005-2009.pdf(file size: 241 KB, MIME type: application/pdf)

HPC File Systems and Scalable I/O: Suggested Research and Development Topics for the fiscal 2005-2009 time frame

DOE Office of Science: Rob Ross ANL, Evan Felix PNL

DOE NNSA: Bill Loewe LLNL, Lee Ward SNL, Gary Grider LANL

DOD: Rob Hill, NSA

Executive Summary

The need for immense and rapidly increasing scale in scientific computation drives the need for rapidly increasing scale in storage for scientific processing. Individual storage devices are rapidly getting denser while bandwidth is not growing at the same pace. In the past several years, Research and Development (R&D) into highly scalable file systems, high level I/O libraries, and I/O middleware was done to provide some solutions to the problems that arise from massively parallel storage. This document primarily concentrates on file systems and I/O middleware since high level I/O libraries have been addressed in many Data Management discussions and calls for R&D. The purpose of this document is to present areas of needed research and development in the HPC file systems and scalable I/O area which should be pursued by the government.

In the last five years, supercomputers with thousands of nodes and over ten thousand processors have been deployed. Additionally, thousands of small clusters have been deployed worldwide. File systems and I/O has come a long way since the simple cross mounted NFS that was used with early clusters. File systems and I/O middleware have been developed and deployed for these supercomputers and clusters systems which have enabled bandwidths beyond ten gigabytes/sec and metadata performance beyond one thousand file inserts/sec. It is now possible to scale bandwidth, and there is competition in the scalable global parallel file systems market space through products that span the gamut from completely proprietary to open source. I/O Middleware is maturing which is enabling many applications to get good performance from the underlying file systems. Recently several new I/O R&D efforts have begun to try to address future needs in HPC file systems and I/O middlware. Efforts in the areas of:

  • Relaxation of POSIX Semantics for parallelism
  • Scalable metadata operations in a single directory
  • NFSv4 security and the pNFS effort to allow NFS to get more native file system

performance through separation of data and control which enables parallelism

  • I/O Middleware enhancements to enable dealing with small, overlapped, and

unaligned I/O

  • Tighter integration between high level I/O libraries and I/O middlware
  • Initial autonomic storage management
  • Sharing of global parallel file systems between multiple clusters
  • Initial active storage research

In the near future, sites will deploy supercomputers with tens of thousands or even hundreds of thousands of processors. Immense bandwidth, metadata, security, and management needs are emerging. Work flows for efficient complex science will begin to approach the exabyte range, and the ability to handle a more varied I/O workload including small to extremely large I/O operations, extremely high metadata activities, and multiple simultaneous workloads will be required. File systems will be so large that complete checks or rebuilds, entire tree walks, and other large operations will not be able HPC File Systems and Scalable I/O: Suggested Research and Development Topics for the fiscal 2005-2009 time frame to be contemplated. Management of these large storage systems will become increasingly difficult. To meet the demands posed by the future HPC environments, investment in R&D and standards work need to be undertaken. The following are key areas to consider investment in:

  • Scaling of metadata , security, reliability, availability, and management to deal

with the enormity of future systems

  • Enhancements in the POSIX I/O API and I/O middleware in the areas of ordering,

coherence, alternate metadata operations, shared file descriptors, locking schemes, and portability of hinting for layouts and other information.

  • Support for machine architectures which do not have full operating systems on

each node and exploitation of hierarchies of node types

  • Additional work in active storage concepts, including use in the LAN/SAN, WAN,

application integration, and object archives

  • Continued support for development and standards work on NFSv4 and pNFS
  • Tracing, Simulation, and benchmarking, including application realistic

benchmarking and simulation

  • New metadata layouts other than tree based directories

More focused and complete government investment needs to be made in the file systems and I/O middlware area of HPC, given its importance and its lack of sufficient funding levels in the past, compared to other elements of HPC. Scalable I/O is perhaps the most overlooked area of HPC R&D, and given the information generating capabilities being installed and contemplated, it is a mistake to continue to neglect this area of HPC. Many areas in need of new and continued investment in R&D and standardization in this crucial HPC I/O area have been summarized in this document.

File history

Click on a date/time to view the file as it appeared at that time.

current09:16, 28 December 2012 (241 KB)Webmaster (talk | contribs) Category:HEC

There are no pages that link to this file.