Kangaroo


Note: Kangaroo is no longer under development.
This page remains for historical interest only.

Kangaroo is a wide-area data-movement system.

A collection of Kangaroo servers can be used to provide robust background data movement for remotely executing programs. Applications write to nearby Kangaroo servers at disk speeds, and rely on background processes to guide it home while they attend to other matters. Kangaroo keeps working, even under hostile conditions.

Kangaroo does not provide low-latency single-file transfers. It does provide high-throughput data movement by overlapping CPU and I/O tasks over a wide area. We measure success in terms of end-to-end application performance.

With the help of the Pluggable File System, applications perceive Kangaroo to be a mere file system. No special privileges are needed to install and use this feature.

Kangaroo is an ongoing research project pursued by Douglas Thain, Jim Basney, Se-Chang Son, and Miron Livny and other members of the Condor Team at the University of Wisconsin-Madison. We would be happy to hear of successes, bugs, questions, and ideas.

Getting Started

  • Introduction (under construction)
  • Technical Manual
  • Software

  • Kangaroo 0_10 for Linux 2.2 x86 (RedHat 6.2)
  • Kangaroo 0_10 for Linux 2.4 x86 (RedHat 7.2)
  • Kangaroo 0_10 for Solaris 5.7 SPARC
  • Kangaroo 0_10 Source
  • Papers

  • Douglas Thain, Jim Basney, Se-Chang Son, and Miron Livny, "The Kangaroo Approach to Data Movement on the Grid", (Postscript) (PDF) in Proceedings of the Tenth IEEE Symposium on High Performance Distributed Computing, San Francisco, California, August 7-9, 2001.
  • Lectures

  • The Kangaroo Approach to Data Movement on the Grid, given at HPDC 10, August 2001.
  • Reliable Data Transfer on the Grid, given at Paradyn/Condor week 2001.
  • The Kangaroo Approach to Data Movement on the Grid, given at the UW OS/Network seminar, March 2001.