Programming Grid Applications with GRID Superscalar

Abstract

The aim of GRID superscalar is to reduce the development complexity of Grid applications to the minimum, in such a way that writing an application for a computational Grid may be as easy as writing a sequential application.Our assumption is that Grid applications would be in a lot of cases composed of tasks, most of them repetitive. The granularity of these tasks will be of the level of simulations or programs, and the data objects will be files. GRID superscalar allows application developers to write their application in a sequential fashion. The requirements to run that sequential application in a computational Grid are the specification of the interface of the tasks that should be run in the Grid, and, at some points, calls to the GRID superscalar interface functions and link with the run-time library.GRID superscalar provides an underlying run-time that is able to detect the inherent parallelism of the sequential application and performs concurrent task submission. In addition to a data-dependence analysis based on those input/output task parameters which are files, techniques such as file renaming and file locality are applied to increase the application performance. This paper presents the current GRID superscalar prototype based on Globus Toolkit 2.x, together with examples and performance evaluation of some benchmarks.

Predictive Runtime Code Scheduling for Heterogeneous Architectures

“…However, in the case of distributed systems, the scheduling is burdened with many more complexities such as interprocessor communication (typically done across some kind of external network), distributed management and variant amount of computing resources (some resources may disappear suddenly, whereas others can turn up). [18,3] Most of the studies related to heterogeneous distributed systems propose new scheduling algorithms in order to improve the performance through job and resource allocation. They mathematically represent a program as a graph where every node is a task which can be mapped on a PE.…”

Integration of GRID Superscalar and GridWay Metascheduler with the DRMAA OGF Standard

“…The application is developed on top of the integration of two powerful grid tools: GRID superscalar (GRIDSs) [1], which provides a grid-unaware programming environment and GridWay [2], a metascheduler which provides reliable execution in heterogeneous Grids. The integration of both tools have been performed through the OGF standard DRMAA [3,4], which has also become the first OGF recommendation.…”

Federated clusters using the transparent remote Execution (TREx) environment

“…In this institutional network, only a few computers are used for scientific computing (large-scale simulations and complex scientific modeling) requiring more resources than what a personal computer provides. These workstations that need more resources for intensive computational programs are considered engineering __________________________________________ 1 TREx experiments and results were obtained during an internship at France Telecom R&D, 38-40 rue du General Leclerc Issy les Moulineaux, France workstations (EW). Normally, the number of clerical stations is greater than the number of engineering workstations.…”

“…Three other well known distributed systems are CORBA [10,11], JINI [10,12] and GRID SuperScalar [1]. In these systems, the clients register their services to a look-up server (broker).…”