|
|
 |
 April
21, 2000
TANet2
Day
Technical
Issues and Applications of the Next Generation Internet
Computational and Data Grids:
Distributed High-Performance Computing
and Large-Scale Data Management for Science and Engineering
William
E. Johnston
Lawrence Berkeley National Laboratory and NASA Ames Research Center
We
use the term "Grid" to refer to the services and infrastructure needed
to support distributed, high performance computing and data handling
systems that incorporate geographically and organizationally dispersed,
heterogeneous resources that are persistent and supported.
Several science and engineering scenarios define the type of Grids
that we are working on in NASA's Information Power Grid and DOE's
Science Grid:
- multi-disciplinary
science and engineering design computational modeling
- large-scale
scientific data analysis
- real-time
data analysis, e.g. for on-line scientific instruments
- coupling
instruments/experiments and computational models
- generation
and management of large, complex data archives
These scenarios
all require some sort of resource aggregation and/or advance scheduling
of resources, and much of the Grid technology is targeted at these
advanced capabilities. However, the biggest near-term impact of the
Grid may be a combination of a service delivery model that promotes
remote use of single resources, and an implementation that provide
a uniform interface to many different resources.
The resulting Grid architecture includes:
- Tools
for constructing collaborative, application oriented Problem Solving
Environments / Frameworks (the primary user interfaces for Grids)
- Programming
environments, tools, and services providing a variety of approaches
for building applications that use aggregated computing and storage
resources, and federated data sources
- A comprehensive
and consistent set of location independent tools and services
for accessing and managing and providing uniform access to dynamic
collections of widely distributed resources: heterogeneous computing
systems, storage systems, real-time data sources and instruments,
human collaborators, and communications systems
- Operational
infrastructure including management tools for distributed systems
and distributed resources, user services, accounting and auditing,
strong and location independent user authentication and authorization,
and overall system security services
The vision for science computation and data Grids is that they will
provide significant new capabilities to scientists and engineers by
facilitating routine construction of information based problem solving
environments / frameworks. Such Grids will knit together widely distributed
computing, data, instrument, and human resources into just-in-time
systems that can address complex and large-scale computing and data
analysis problems.
This talk will discuss the motivation and architecture of science
Grids, and implementation progress of NASA's IPG: A prototype production
Grid.
|
April
21, 2000 
