1989RIT Scholar Works (Rochester Institute of Technology)Open access

The DFS distributed file system: Design and implementation

Ananth K. Rao

Open full text 0 citations

Abstract

This thesis describes the design of an operating system independent distributed file system (DFS) and details the implementation, on a cooperating set of server computers interconnected by means of a communications network. The DFS provides the mechanism by which the file systems of these physically dispersed units are integrated into a single logical unit. Users and application programs thus have the illusion that their files are on a single computer system, even though in reality they may be physically distributed around the network. This location transparency frees users from having to remember details such as the current location of the file and also affords considerable mobility, allowing access to files from any workstation. In addition, automatic storage replication and an atomic transaction mechanism provides high reliability and improved availability in the distributed environment in the face of site failure.

Open-access reader

About this research paper

What this paper is about

This thesis describes the design of an operating system independent distributed file system (DFS) and details the implementation, on a cooperating set of server computers interconnected by means of a communications network. The DFS provides the mechanism by which the file systems of these physically dispersed units are integrated into a single logical unit. Users and application programs thus have the illusion that their files are on a single computer system, even though in reality they may be physically distributed around the network. This location transparency frees users from having to remember details such as the current location of the file and also affords considerable mobility, allowing access to files from any workstation. In addition, automatic storage replication and an atomic transaction mechanism provides high reliability and improved availability in the distributed environment in the face of site failure.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

This thesis describes the design of an operating system independent distributed file system (DFS) and details the implementation, on a cooperating set of server computers interconnected by means of a communications network. The DFS provides the mechanism by which the file systems of these physically dispersed units are integrated into a single logical unit. Users and application programs thus have the illusion that their files are on a single computer system, even though in reality they may be physically distributed around the network. This location transparency frees users from having to remember details such as the current location of the file and also affords considerable mobility, allowing access to files from any workstation. In addition, automatic storage replication and an atomic transaction mechanism provides high reliability and improved availability in the distributed environment in the face of site failure.

Key concepts: Distributed File System, Self-certifying File System, Computer science, Network File System, File system, Operating system, Unix file types, SSH File Transfer Protocol

Related papers

Back to paper searchBrowse research topicsOriginal source
The DFS distributed file system: Design and implementation — Research Paper | ScholarLens