From Java CoG Kit

Gregor von Laszewski, Keith Jackson

Problem Statement

The computing demand of many state-of-the-art scientific applications, such as climate modeling, astrophysics, high energy physics, structural biology, chemistry, and tele-immersive engineering, will require the coordinated use of numerous distributed and heterogeneous components, including advanced networks, computers, storage devices, display devices, and scientific instruments. Such a national collaborative Grid infrastructure is currently being developed and supported by DOE, NSF, and NASA. Developing advanced scientific applications for these emerging national-scale Computational Grid infrastructures is still a difficult task. Although elementary Grid services are available that assist the application developers in authentication, remote access to computers, resource management, and infrastructure discovery, these services often are not compatible with the commodity technologies and frameworks used by application scientists today, or they are too complex to be used by computational scientists who are not experts in Grid programming. A higher level of abstraction is demanded that provides advanced Grid services within frameworks that are, and will be, used as part of the scientific problem-solving process.

We will develop and implement Commodity Grid (CoG) Kits that will allow the application programmer or middleware developer to readily make use of Grid services from a higher-level framework. We will focus this effort on the development of two CoG Kits; one for Java and the other for Python. These kits will allow for easier and more rapid application development. These kits will allow for easier and more rapid application development. They also will encourage collaborative code reuse and avoid the duplication of effort among problem solving environments, science portals, Grid middleware, and collaboratory pilots.

Documents

• Executive summary (old) PDF
• 2 page summary PDF
• One slide summary PDF
• Ten slide summary PDF
• Poster PDF
• Selected References PDF
• Tutorial: Java and Python for Grid Computing, Gregor von Laszewski, Keith Jackson, Argonne Preprint ANL/MCS-P882-050, PDF
• Status Report PDF

Links

• Java CoG Kit: http://www.globus.org/cog/
• Python CoG Kit: http://www-itg.lbl.gov/gtg/projects/pyGlobus/

Management

The project is managed by Gregor von Laszewski and Keith Jackson. Regular meetings are held via collaborative tools. Face-to-face meetings are held at time of major other meetings and conferences as to keep the cost for traveling down. Written reports help to monitor the progress and goals of the project. Interaction with SciDAC projects and other projects verify the applicability of the CoG Kits within the project scope.

Timeline

• Year 1, 2002::
  • Identify the Requirements of the community
  • Provide access to basic Grid services in Python and Java
  • Release the developed software to the community

• Year 2, 2003::
  • Provide success stories for the use of the CoG Kit with application and Grid service developers
  • Provide advanced Grid services
  • Interact with application development teams

• Year 3, 2004::
  • Provide extensive tutorials and documentation for an open source development
  • Provide a community release of the software
  • Verify the development of independent maintained CoG Kits

• Year 4, 2005::
  • Provide a Grid Worklfow Stystem
  • Develop Grid Abstractions

• Year 5, 2006::
  • Funding provided thro NSF DDDAS

Contact

Gregor von Laszewski   
Argonne National Laboratory  
9700 S. Cass Ave             
Argonne, IL 60439            
phone: (630) 252-0472        
fax: (630) 252-5986          
gregor@mcs.anl.gov           

Keith Jackson          
Lawrence Berkeley National Laboratory
1 Cyclotron Rd. 50B-2239     
Berkeley, CA 94720           
phone: (510) 486-4401        
fax: (510) 486-6363          
KRJackson@lbl.gov