Grid Programming Models: Current Tools, Issues and Directions

Abstract

Grid programming must manage computing environments that are inherently parallel, distributed, heterogeneous and dynamic, both in terms of the resources involved and their performance. Furthermore, grid applications will want to dynamically and flexibly compose resources and services across that dynamic environments. While it may be possible to build grid applications using established programming tools, they are not particularly well-suited to effectively manage flexible composition or deal with heterogeneous hierarchies of machines, data and networks with heterogeneous performance. This chapter discusses issues, properties and capabilities of grid programming models and tools to support efficient grid programs and their effective development. The main issues are outlined and then current programming paradigms and tools are surveyed, examining their suitability for grid programming. Clearly no one tool will address all requirements in all situations. However, paradigms and tools that can incorporate and provide the widest possible support for grid programming will come to dominant. Advanced programming support techniques are analyzed discussing possibilities for their effective implementation on grid environments.

Efficient Parallel Application Execution on Opportunistic Desktop Grids

“…While it is largely acknowledged that no one-size-fits-all solution exists when it comes to programming models, one can argue that some programming models are best suited for particular kinds of problems than others (Lee & Talia, 2003). Considering that grid computing environments can be used to run different kinds of applications in different domains, such as e-science, finance, numeric simulation and, more generally, virtual organizations, it follows that having a variety of programming model to choose from may be an important factor.…”

“…Considering that grid computing environments can be used to run different kinds of applications in different domains, such as e-science, finance, numeric simulation and, more generally, virtual organizations, it follows that having a variety of programming model to choose from may be an important factor. A number of well-known programming models have been investigated for grid computing, including, but not limited to, remote procedure calls (as in RPC and RMI), tuple spaces, publish-subscribe, message passing, and Web services, as well as enhancements of such models with non-functional properties such as fault tolerance, dependability and security (Lee & Talia, 2003). Note that a main emphasis of programming models for grid computing is on communication abstractions, which is due to the fact that interaction among distributed application components is a key issue for grid applications.…”

Towards Distributed Petascale Computing

“…For tightly coupled HPC applications this is not so clear, as common wisdom is that running a tightly coupled parallel application in a computational grid (in other words, a parallel job actually running on several parallel machines that communicate with each other in a Grid) is of no general use because of the large overheads that will be induced by communications between computing elements (see e.g. [17]). However, in our opinion this certainly is a viable option, provided the granularity of the computation is large enough to overcome the admittedly large communication latencies that exist between compute elements in a Grid.…”

On Grid Programming and MATLAB*G

“…The main goal of grid programming is the study of programming models, tools and methods that support the effective development of portable and high-performance algorithms and applications on grid environments [20]. Grid programming will require capabilities and properties beyond that of simple sequential programming or even parallel and distributed programming.…”