A Holographic File System for a Multicomputer with Many Disk Nodes

Abstract

Future computing systems may involve thousands of networked general- purpose computers, without shared memory or shared devices. The operating system for such a configuration must be completely distributed, and it must tolerate the random disappearance and reappearance of nodes, possibly with obsolete control information and/or data. Traditional file systems are not equipped to satisfy these requirements. A holographic file system (HFS) is presented for concurrent data retrieval in a computer system with a large number of disks (although it is probable that most nodes will be diskless). In such a file system it is possible to operate on a file by simultaneously utilizing many (or all) of the disks, since the file system is organized to take advantage of the multiplicity of equipment, rather than limiting access to a single disk for each file, as in most existing file systems. The main advantages of the HFS are a speed -up in data retrieval related to the number of disks, and improved availability by allowing access to parts of a file even when other parts of that file are not accessible. Keywords: Distributed file system; Holographic file system; Interleaved file system; File servers; File system organization; Diskless workstations; Distributed systems.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 1988
Accession Number
ADA201875

Entities

People

  • Amnon Barak
  • Bernard A. Galler
  • Yaron Farber

Organizations

  • Hebrew University of Jerusalem

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Access Time
  • Algorithms
  • Communication Networks
  • Computer Programs
  • Computers
  • Computing System Architectures
  • Congestion
  • Databases
  • Directories
  • Distributed Computing
  • Operating Systems
  • Probability
  • Random Variables
  • Servers (Computer Hardware)
  • Simulations
  • Time Intervals
  • Trees (Data Structures)

Fields of Study

  • Computer science

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Parallel and Distributed Computing.