The Experts below are selected from a list of 237 Experts worldwide ranked by ideXlab platform

David G. Andersen - One of the best experts on this subject based on the ideXlab platform.

  • using vector interfaces to deliver millions of iops from a networked key value Storage server
    Symposium on Cloud Computing, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

  • SoCC - Using vector interfaces to deliver millions of IOPS from a networked key-value Storage server
    Proceedings of the Third ACM Symposium on Cloud Computing - SoCC '12, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

Vijay K. Vasudevan - One of the best experts on this subject based on the ideXlab platform.

  • using vector interfaces to deliver millions of iops from a networked key value Storage server
    Symposium on Cloud Computing, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

  • SoCC - Using vector interfaces to deliver millions of IOPS from a networked key-value Storage server
    Proceedings of the Third ACM Symposium on Cloud Computing - SoCC '12, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

Michael Kaminsky - One of the best experts on this subject based on the ideXlab platform.

  • using vector interfaces to deliver millions of iops from a networked key value Storage server
    Symposium on Cloud Computing, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

  • SoCC - Using vector interfaces to deliver millions of IOPS from a networked key-value Storage server
    Proceedings of the Third ACM Symposium on Cloud Computing - SoCC '12, 2012
    Co-Authors: Vijay K. Vasudevan, Michael Kaminsky, David G. Andersen
    Abstract:

    The performance of non-volatile memories (NVM) has grown by a factor of 100 during the last several years: Flash devices today are capable of over 1 million I/Os per second. Unfortunately, this incredible growth has put strain on Software Storage systems looking to extract their full potential. To address this increasing Software-I/O gap, we propose using vector interfaces in high-performance networked systems. Vector interfaces organize requests and computation in a distributed system into collections of similar but independent units of work, thereby providing opportunities to amortize and eliminate the redundant work common in many high-performance systems. By integrating vector interfaces into Storage and RPC components, we demonstrate that a single key-value Storage server can provide 1.6 million requests per second with a median latency below one millisecond, over fourteen times greater than the same Software absent the use of vector interfaces. We show that pervasively applying vector interfaces is necessary to achieve this potential and describe how to compose these interfaces together to ensure that vectors of work are propagated throughout a distributed system.

L. G. Blasso - One of the best experts on this subject based on the ideXlab platform.

  • NSSDC Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications, volume 1
    1992
    Co-Authors: Benjamin Kobler, P. C. Hariharan, L. G. Blasso
    Abstract:

    Papers and viewgraphs from the conference are presented. This conference served as a broad forum for the discussion of a number of important issues in the field of mass Storage systems. Topics include magnetic disk and tape technologies, optical disks and tape, Software Storage and file management systems, and experiences with the use of a large, distributed Storage system. The technical presentations describe, among other things, integrated mass Storage systems that are expected to be available commercially. Also included is a series of presentations from Federal Government organizations and research institutions covering their mass Storage requirements for the 1990's.

  • NSSDC Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications, volume 2
    1992
    Co-Authors: Ben Kobler, P. C. Hariharan, L. G. Blasso
    Abstract:

    This report contains copies of nearly all of the technical papers and viewgraphs presented at the NSSDC Conference on Mass Storage Systems and Technologies for Space and Earth Science Application. This conference served as a broad forum for the discussion of a number of important issues in the field of mass Storage systems. Topics include the following: magnetic disk and tape technologies; optical disk and tape; Software Storage and file management systems; and experiences with the use of a large, distributed Storage system. The technical presentations describe, among other things, integrated mass Storage systems that are expected to be available commercially. Also included is a series of presentations from Federal Government organizations and research institutions covering their mass Storage requirements for the 1990's. Separate abstracts were prepared for papers in this report.

  • NSSDC Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications, volume 3
    1992
    Co-Authors: Ben Kobler, P. C. Hariharan, L. G. Blasso
    Abstract:

    This report contains copies of nearly all of the technical papers and viewgraphs presented at the National Space Science Data Center (NSSDC) Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications. This conference served as a broad forum for the discussion of a number of important issues in the field of mass Storage systems. Topics include magnetic disk and tape technologies, optical disk and tape, Software Storage and file management systems, and experiences with the use of a large, distributed Storage system. The technical presentations describe, among other things, integrated mass Storage systems that are expected to be available commercially. Also included is a series of presentations from Federal Government organizations and research institutions covering their mass Storage requirements for the 1990s. Separate abstracts have been prepared for papers in this report.

  • Proceedings of the NSSDC Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications
    1991
    Co-Authors: Kim Blackwell, L. G. Blasso, Ann Lipscomb
    Abstract:

    The proceedings of the National Space Science Data Center Conference on Mass Storage Systems and Technologies for Space and Earth Science Applications held July 23 through 25, 1991 at the NASA/Goddard Space Flight Center are presented. The program includes a keynote address, invited technical papers, and selected technical presentations to provide a broad forum for the discussion of a number of important issues in the field of mass Storage systems. Topics include magnetic disk and tape technologies, optical disk and tape, Software Storage and file management systems, and experiences with the use of a large, distributed Storage system. The technical presentations describe integrated mass Storage systems that are expected to be available commercially. Also included is a series of presentations from Federal Government organizations and research institutions covering their mass Storage requirements for the 1990's.

Mérouane Debbah - One of the best experts on this subject based on the ideXlab platform.

  • Cloud Storage for Small Cell Networks
    2012 IEEE 1st International Conference on Cloud Networking (CLOUDNET), 2012
    Co-Authors: Ejder Baştuğ, Jean-louis Guénégo, Mérouane Debbah
    Abstract:

    The Massive dense deployment of Small Cell Networks (SCNs) is a promising way of increasing capacity. Interestingly, by having such a huge amount of user devices as well as small cells deployed in indoor or outdoor areas, one can take benefit of such distributed network for Storage purposes. Hence, Depending on the deployed scenario, these Storage units can also be used for content caching to relieve the backhaul constraints and increase the peak rate. In this work, by extending concepts of cloud Storage to SCNs, we discuss the theoretical challenges in order to embed small cells with Storage capabilities. We also briefly introduce Open Cloud Protocol (OCP) as a unified Software Storage framework.

  • CLOUDNET - Cloud Storage for Small Cell Networks
    2012 IEEE 1st International Conference on Cloud Networking (CLOUDNET), 2012
    Co-Authors: Ejder Baştuğ, Jean-louis Guénégo, Mérouane Debbah
    Abstract:

    The Massive dense deployment of Small Cell Networks (SCNs) is a promising way of increasing capacity. Interestingly, by having such a huge amount of user devices as well as small cells deployed in indoor or outdoor areas, one can take benefit of such distributed network for Storage purposes. Hence, Depending on the deployed scenario, these Storage units can also be used for content caching to relieve the backhaul constraints and increase the peak rate. In this work, by extending concepts of cloud Storage to SCNs, we discuss the theoretical challenges in order to embed small cells with Storage capabilities. We also briefly introduce Open Cloud Protocol (OCP) as a unified Software Storage framework.