OptimalGrid
A research prototype of grid-enabled middleware designed to hide complexities of partitioning, distributing, and load balancing.
Date Posted: April 16, 2003
|
|
 |
|
|
Update: September 12, 2005
This release includes a new easily-used provisioning and launch facility, updated load balancing, and the first release of the code for running Quake II on a grid as a massively multiplayer game.
What is OptimalGrid?
OptimalGrid is a research prototype of a grid-enabled collaboration framework, sophisticated management infrastructure, and problem-solving environment for grid computing that is designed to hide complexities of partitioning, distributing, and load balancing.
The goal of OptimalGrid is to simplify the creation and management of connected parallel applications on the grid by including important autonomic grid functionality in a middleware layer. Ideally, developers should be able to create grid-enabled parallel applications without themselves becoming experts in grid or high-performance computing. Grid applications should be able to automatically reconfigure themselves in response to dynamic changes in the grid environment.
The new release of OptimalGrid includes numerous refinements of the OptimalGrid system and includes two interesting demonstations of OptimalGrid's abilities. The first is the Eden model of bacterial growth running in both a stand-alone mode (simulating a grid on one machine) and a distributed mode on multiple grid nodes. The second is the Quake II game from Id Software transformed in a massively multiplayer engine running on a grid.
This technology is Linux®-compatible.
How does it work? OptimalGrid is not a toolkit; it is self-contained middleware. It can either use the underlying grid node management software or provide its own basic grid infrastructure as needed. Developers would write applications as plug-ins, which OptimalGrid then distributes and runs on a grid.
OptimalGrid will run on almost any grid infrastructure and requires only that a Java Run-time Environment be installed on networked machines. It is designed to optimize performance in order to make the most of an existing grid infrastructure.
For inquiries regarding commercial licensing of OptimalGrid, please e-mail Leigh Cagan, Manager of the Business Development Department of Almaden Research Center.
|
|
|
| | About the technology author(s):
Glenn Deen is the leader of the OptimalGrid project at the IBM Almaden Research Center, San Jose, California. Since joining IBM in 1989, he has been involved in security architecture development and distributed computing solutions. His current interests include grid computing, large distributed simulations, and electronic healthcare systems.
James H. Kaufman is a research staff member in the Distributed & Cluster Systems Department at the IBM Almaden Research Center. He received a Ph.D. in physics from U.C.S.B. and a B.A. in physics from Cornell University. Dr. Kaufman has made contributions to several fields of research at IBM, and he is a fellow of the American Physical Society. His current research interests include distributed computing, simulation and modeling, and grid middleware.
Tobin (Toby) J. Lehman joined the IBM Almaden Research Center in 1986, shortly after finishing his Ph.D. at the University of Wisconsin-Madison. His research interests include server-based back-up systems, object-relational database systems, large object management, memory-resident database systems, Tuplespace systems, and computing grids. Dr. Lehman is currently working on an autonomic grid computing infrastructure for solving very large connected problems on a heterogeneous collection of Internet machines.
Iris Eiron is a researcher at the IBM Almaden Research Center. She joined IBM in January 1998 after receiving her M.Sc. in computer science from the Technion, the Israeli Institute of Technology. Ms. Eiron worked for the IBM Israeli Research Lab for three years before joining the Almaden Research Center in December 2000. Her current interests include development and implementation of a national health care infrastructure.
John Thomas is a Java developer for IBM. He was formerly one of the lead programmers for the IBM Almaden TSpaces project. Mr. Thomas is currently a member of the OptimalGrid Project at the Almaden Research Center.
These engineers can all be reached through e-mail.
|
|
|
For platform(s):
Windows NT, Linux, Windows 2000, Windows, Java
|
 |
For topics:
Autonomic computing, games, Life sciences, load balancing, optimization, partitioning, performance, Grid computing, Middleware
|
|
| |
