Copernicus: A Scalable, High-Performance Semantic File System

Published as Storage Systems Research Center Technical Report UCSC-SSRC-09-06.

Abstract

Hierarchical file systems do not effectively meet the needs of users
at the petabyte-scale. Users need dynamic, search-based file access in
order to properly manage and use their growing sea of data. This paper
presents the design of Copernicus, a new scalable, semantic file
system that provides a searchable namespace for billions of
files. Instead of augmenting a traditional file system with a
search index, Copernicus uses a dynamic, graph-based file system
design that indexes file attributes and relationships to provide
scalable search and navigation of files.

Available for download:

Bibtex entry

@techreport{
  author       = {Andrew Leung and Aleatha Parker-Wood and Ethan L. Miller},
  title        = {Copernicus: A Scalable, High-Performance Semantic File System},
  institution  = {University of California, Santa Cruz},
  number       = {UCSC-SSRC-09-06},
  month        = oct,
  year         = {2009},
}