Using a network ring buffer for science and outreach

Paul Hubbard
Jason Hanley
Terry Weymouth
Matthew Miller

The general architecture, base case:
Architecture diagram
Our supercool integrated TiVo-style client, RDV:
RDV data viewer showing an experiment

Submitted to (and rejected by) supercomputing 2005, this paper is a needs-improvement explanation of our use of a network ring buffer in distributed research.

Since its doing no good laying about my hard drive, I decided to post it to the web and let the magic of google do the rest. If this is of interest, read and enjoy! As with a previous paper, I've also posted the reviewers' comments.

All of the code we've written is open source, so please mail me if you want it. We're converting from Perforce to Subversion now; once that's complete I'll post the svn URL here.

Keywords: Java, ring bufffer, distributed experiments, telepresence, earthquake engineering.

The paper itself, PDF, 1.3MB
The reviews (painful to read!), PDF, 112KB

Back to home page