The design and implementation of the GridLab information service

International Journal of High Performance Computing Applications, 2003

Grid technology is widely emerging. Still, there is an eminent shortage of real Grid users, mostly due to the lack of a "critical mass" of widely deployed and reliable higher-level Grid services, tailored to application needs. The GridLab project aims to provide fundamentally new capabilities for applications to exploit the power of Grid computing, thus bridging the gap between application needs and existing Grid middleware. We present an overview of GridLab, a large-scale, EU-funded Grid project spanning over a dozen groups in Europe and the US. We first outline our vision of Grid-empowered applications and then discuss GridLab's general architecture and its Grid Application Toolkit (GAT). We illustrate how applications can be Grid-enabled with the GAT and discuss GridLab's scheduler as an example of GAT services.

The term \Grid" has become common parlance among parallel and distributed computer scientists to denote a middleware infrastructure for wide-area scienti c and engineering computing. The discussion of standards and best practices is ongoing in the Grid Forum community to identify key pieces of the middleware infrastructure. One of the areas under discussion is di erent models for grid information services.

Grid Computing. Wiley InterScience, Hoboken, NJ, S, 2003

Part A of this book, chapters 1 to 5, provides an overview and motivation for Grids. Further chapter 37 is an illuminating discussion from 1992 of Metacomputing-a key early concept on which much of the Grid has been built. Chapter 2 is a short overview of the Grid reprinted from Physics Today [14]. Chapter 3 gives a detailed recent history of the Grid while chapter 4 describes the software environment of the seminal I-WAY experiment at SC95. This conference project challenged participantsincluding for instance the Legion activity of chapter 10-to demonstrate Grid-like applications on an OC-3 backbone. Globus [6] grew out of the software needed to support these 60 applications at 17 sites; the human intensive scheduling and security used by I-WAY showed the way to today's powerful approaches. Many of these applications employed visualization including CAVE virtual reality stations as demonstrated in the early Metacomputing work of chapter 37. Chapter 5 brings us to 2002 and describes the experience building Globus-based grids for NASA and DoE. Metacomputing and hence the Grid was born in the High Performance Computing and Communication (HPCC) activities of the 1980's and 1990's. In particular the multiagency Grand (application) Challenges brought critical issues to the fore. We realized the importance of coupling programs together to solve multidisciplinary applications. Here we see beginnings of "Science as a team sport" elaborated in section 9. We saw applications like that of figure 2 that linked instruments, visualization, computing and data [3]-often in a pipeline-and were one of the first important classes of successful Grid applications. HPCC of course built on the growing use of parallel computing and software infrastructure like PVM, MPI, HPF, and OpenMP to support scalable applications. Initially it was thought that the Grid would be most useful in extending these parallel computing approaches from tightly coupled clusters to geographically distributed systems. However more important has been the integration of loosely coupled system-each component of which might be running in parallel on a low-latency parallel machine. The critical Grid task of managing these heterogeneous components as we scale the size of distributed systems replaces that of the tight synchronization of the typically identical (in program but not data as in SPMD-single program multiple data-model) parts of a domain decomposed parallel application. Networking was injected into the Grid-initially with initiatives like the Gigabit testbed program [15] with its projects-Aurora, Blanca, Casa, Nectar, and Vistanet-and dual goals: to investigate potential testbed network architectures, and to explore their usefulness for end-users. Computational Grids got their name from analogies with other national or global infrastructure systems [16]. They share properties with Electric Power Grids; both are ubiquitous; in both cases one does not need to know source of (electric or computing) power (transformer or generator, PC or supercomputer) and the supplying organization. Computational Grids also have differences-they have a wider performance spectrum, more and more heterogeneous services, complex access and security issues, and complex socio-political factors. As we move from distributed to Grid computing, we are moving from largely addressing geographical separation to a focus on the integration and management of software. We are developing a new software engineering as the Grid Web services described starting in chapter 7, provide a component model applicable to any software development project. Interesting features include intrinsic self-documentation of such

A scientist studying proteins logs into a computer and uses an entire network of computers to analyze data. A businessman accesses his company's network through a PDA in order to forecast the future of a particular stock. An Army official accesses and coordinates computer resources on three different military networks to formulate a battle strategy. All of these scenarios have one thing in common: They rely on a concept called grid computing. At its most basic level, grid computing is a computer network in which each computer's resources are shared with every other computer in the system. Processing power, memory and data storage are all community resources that authorized users can tap into and leverage for specific tasks. A grid computing system can be as simple as a collection of similar computers running on the same operating system or as complex as inter-networked systems comprised of every computer platform you can think of.

2002

Grid technology is widely emerging. Still, there is an eminent shortage of real Grid users, due to the absence of two important catalysts: First, a widely accepted vision on how applications can substantially benefit from Grids, and second a toolkit of higher-level Grid services, tailored to application needs. The GridLab project aims to provide fundamentally new capabilities for applications to exploit the power of Grid computing, thus bridging the gap between application needs and existing Grid middleware. We present an overview of GridLab, a largescale, EU-funded Grid project spanning over a dozen groups in Europe and the US. We first outline our vision of Grid-empowered applications and then discuss GridLab’s general architecture.

2003

We describe Peer-to-Peer Grids built around Integration of technologies from the peer-to-peer and Grid fields. We focus on the role of Web services linked by a powerful event service using uniform XML interfaces and application level routing. We describe how a rich synchronous and asynchronous collaboration environment can support virtual communities built on top of such infrastructure. Universal access mechanisms are discussed.