A Case for RAMClouds: Scalable High-Performance Storage Entirely in DRAM
, Parag Agrawal, David Erickson, Christos Kozyrakis, Jacob Leverich, David Mazieres, Subhasish Mitra, Aravind Narayanan, Diego Ongaro, Guru Parulkar, Mendel Rosenblum, Stephen Rumble, Eric Stratmann, Ryan Stutsman

Citation
, Parag Agrawal, David Erickson, Christos Kozyrakis, Jacob Leverich, David Mazieres, Subhasish Mitra, Aravind Narayanan, Diego Ongaro, Guru Parulkar, Mendel Rosenblum, Stephen Rumble, Eric Stratmann, Ryan Stutsman. "A Case for RAMClouds: Scalable High-Performance Storage Entirely in DRAM". Communications of the ACM, pp. TBD, 2010; This paper has been accepted for publication, but the publication date has not yet been determined.

Abstract
Disk-oriented approaches to online storage are becoming increasingly problematic: they do not scale gracefully to meet the needs of large-scale Web applications, and improvements in disk capacity have far out-stripped improvements in access latency and bandwidth. This article argues for a new approach to datacenter storage called RAMCloud, where all information is kept in DRAM at all times and large-scale systems are created by aggregating the main memories of thousands of commodity servers. We believe that RAMClouds can provide durable and available storage with 100-1000x the throughput of disk-based systems and 100-1000x lower access latency. This combination of low latency and large scale will enable a new breed of data-intensive applications.

Electronic downloads

Internal. This publication has been marked by the author for MuSyC-only distribution, so electronic downloads are not available without logging in.

Citation formats

• HTML

   , Parag Agrawal, David Erickson, Christos Kozyrakis, Jacob
  Leverich, David Mazieres, Subhasish Mitra, Aravind
  Narayanan, Diego Ongaro, Guru Parulkar, Mendel Rosenblum,
  Stephen Rumble, Eric Stratmann, Ryan Stutsman. <a
  href="http://www.musyc.org/pubs/87.html" >A
  Case for RAMClouds: Scalable High-Performance Storage
  Entirely in DRAM</a>, <i>Communications of the
  ACM</i>, pp. TBD,  2010; This paper has been accepted
  for publication, but the publication date has not yet been
  determined.

• Plain text

   , Parag Agrawal, David Erickson, Christos Kozyrakis, Jacob
  Leverich, David Mazieres, Subhasish Mitra, Aravind
  Narayanan, Diego Ongaro, Guru Parulkar, Mendel Rosenblum,
  Stephen Rumble, Eric Stratmann, Ryan Stutsman. "A Case
  for RAMClouds: Scalable High-Performance Storage Entirely in
  DRAM". <i>Communications of the ACM</i>,
  pp. TBD,  2010; This paper has been accepted for
  publication, but the publication date has not yet been
  determined.

• BibTeX

  @article{AgrawalEricksonKozyrakisLeverichMazieresMitraNarayanan10_CaseForRAMCloudsScalableHighPerformanceStorageEntirely,
      author = {  and Parag Agrawal and David Erickson and
                Christos Kozyrakis and Jacob Leverich and David
                Mazieres and Subhasish Mitra and Aravind Narayanan
                and Diego Ongaro and Guru Parulkar and Mendel
                Rosenblum and Stephen Rumble and Eric Stratmann
                and Ryan Stutsman},
      title = {A Case for RAMClouds: Scalable High-Performance
                Storage Entirely in DRAM},
      journal = {Communications of the ACM},
      number = {TBD},
      pages = {TBD},
      year = {2010},
      note = {This paper has been accepted for publication, but
                the publication date has not yet been determined.},
      abstract = {Disk-oriented approaches to online storage are
                becoming increasingly problematic: they do not
                scale gracefully to meet the needs of large-scale
                Web applications, and improvements in disk
                capacity have far out-stripped improvements in
                access latency and bandwidth. This article argues
                for a new approach to datacenter storage called
                RAMCloud, where all information is kept in DRAM at
                all times and large-scale systems are created by
                aggregating the main memories of thousands of
                commodity servers. We believe that RAMClouds can
                provide durable and available storage with
                100-1000x the throughput of disk-based systems and
                100-1000x lower access latency. This combination
                of low latency and large scale will enable a new
                breed of data-intensive applications.},
      URL = {http://www.musyc.org/pubs/87.html}
  }
  

Posted by Array on 5 Aug 2010.