NLANR/DAST : Introduction to The Grid

Go to the top of the NLANR/DAST web site

About:

- DAST

- NLANR

- FAQ

- Staff

- Contact DAST

End User Tools and Projects
- NextINet
- Advanced Applications
Database
- DAST Projects/Tools
- Network Performance
and Measurement Tools

End User Support
- Getting Started Guide
- Networking Glossary
- Other Projects/Organizations
- Funding Opportunities

Documents
- Guides/Tutorials
- Papers/Articles
- Presentations
- Reference Books

WebCT Courses
- Tuning Applications

Events
- NLANR/DAST Training
- NLANR Packets Calendar
- Idesk Travel Schedule

News
- Press Releases
- Alliance Data Link
- I2 Newswire Archives

Reports & Statistics
- Monthly Updates and QSRs
- Abilene "Weather Map"
- Web Server Stats

Introduction Grids Globus

Introduction to The Grid
A Little History
Requirements for a Usable Grid
Other Issues

Introduction to The Grid

With the advent of new high-speed research networks like the vBNS and Abilene, linking together high-performance resources is now possible. The Grid merges people, computers, databases, instruments, and other resources in ways that were simply never possible before.

The Grid is already contributing to a variety of research projects, including:

Computer-enhanced instruments, such as this example in microtomography

Collaborative engineering, such as this example creating a collaborative learning environment for young children

Very large-scale simulation, such as this example of vehicle simulation

Browsing of remote datasets

Use of remote software

Data-intensive computing

Large-scale parameter studies

Applications that use the Grid need a relatively easy way to access such distributed resources. The Grid is about bringing resources together, not merely about making more compute cycles available.

A Little History

In the late 1980s the term metacomputing was coined by NCSA Director Larry Smarr. Metacomputing initially meant running computations across multiple machines and was the precursor to truly distributed computing. When developed, the Grid became a superset of metacomputing.

In the early 1990s, Gigabit Testbeds started to be available. These testbeds were research-oriented and helped solve the networking problems needed to get high-speed connections from point A to point B. The research focus was on connection and bandwidth, not necessarily on getting applications to run.

By 1995, the Gigbit Testbeds had shown that a high-speed connection could be established and maintained. The next logical step was to generate "real" applications that could help move the testbed functionality away from networking and toward application development. Those involved realized that some interesting new things could be accomplished if high-speed networking could be used to connect the various resources across the network.

At Supercomputing 95, researchers were solicited to do network computing that involved supercomputers, Immersadesks, high-speed networks, and high-end visualization. The I-WAY demonstration at SC95 connected dozens of centers world-wide via high-speed OC-3 networks in order to run over 60 applications for one week in San Diego. I-WAY was the first demonstration to clearly show there was a whole new kind of application that was suddenly possible, if the resources could be brought together in a coherent network.

A large segment of the research community, including funding agencies, became aware in 1995 of the concept that would eventually become the Grid. Today active programs are working on various elements of Grid development, as Larry Smarr noted in a presentation in early 1999. The various agencies and organizations recognized the potential of bringing these resources together:

NCSA and the San Diego Supercompter Center both decided to participate in the NSF's Partnership for Advanced Computational Infrastucture program (1998).

NASA established the Information Power Grid (1999) that linked together Ames, Glenn (formerly Lewis), and Langley facilities.

DOE established ASCI DISCOM (Advanced Strategic Computing Initiative, DIStributed COMputing effort) (1999) that tied together three weapons labs and their users.

Vice-President Al Gore's office set-up IT², which cuts across all federal agencies to push the high-end level of computing capability.

Requirements for a Useable Grid

For a Grid to be truly useful, it has to provide dependable, consistent, pervasive access to high-end resources. A good analogy is the electrical power grid established in the 1930s:

Dependable
Electricity is always there. When the switch is thrown, the light always comes on. A power outage is an unusual situation in normal, day-to-day life.
Consistent
The uniform interface means that all appliances use the same connector. With minor adaptations, a plug works almost everywhere in the world.
Pervasive
Electrical service is virtually everywhere, throughout the world.

Another example of something having these properties is the Internet Protocol (IP), which is ubiquitous, standardized, and everywhere.

Other Issues

Grid computing requires the ability to work with complex application structures, and combines the aspects of parallel, multimedia, distributed, and collaborative computing, each of which is a full discipline in its own right.

Grid computing takes place in an environment where dynamic, varying resource characteristics in both time and space are the norm, and where there is a need for high and guaranteed "end-to-end" performance, despite heterogeneity and a lack of global control.

Finally, interdomain issues of security, policy, and payment must be taken into account to enable resources from many different institutions to be available on the Grid.

Contact DAST Blank Space Last reviewed: December 31, 1969
NLANR || Applications Support || Engineering Support || Measurement and Network Analysis