XROOTD - A highly scalable architecture for data access

Alvise Dorigo, Peter Elmer, Fabrizio Furano, Andrew Hanushevsky

Research output: Contribution to journalArticlepeer-review

92 Scopus citations

Abstract

When dealing with the concurrent access from a multitude of clients to petabyte-scale data repositories, high performance, fault tolerance, robustness, and scalability are four very important issues. This work describes the architecture and the choices done in designing the xrootd file access system. The first goal of the system was to provide access to over 107 files representing several petabytes of experimental physics data. This work addresses the high demand data access needs of modern physics experiments, such as the BaBar experiment at SLAC, and covers method and tools useful to any other field in which reliability, performance and scalability in data access are a primary issue.

Original languageEnglish (US)
Pages (from-to)348-353
Number of pages6
JournalWSEAS Transactions on Computers
Volume4
Issue number4
StatePublished - Apr 2005

All Science Journal Classification (ASJC) codes

  • General Computer Science

Keywords

  • Fault tolerance
  • Load balancing
  • Peer to peer
  • ROOT
  • Scalability
  • TXNetFile
  • XROOTD

Fingerprint

Dive into the research topics of 'XROOTD - A highly scalable architecture for data access'. Together they form a unique fingerprint.

Cite this