An enabling framework for master-worker applications on the Computational Grid

Abstract. This article presents Grid'BnB, a parallel branch and bound framework for grids. Branch and bound (B&B) algorithms find optimal solutions of search problems and NP-hard optimization problems. Grid'BnB is a Java framework that helps programmers to distribute problems over grids by hiding distribution issues. It is built over a masterworker approach and provides a transparent communication system among tasks. This work also introduces a new mechanism to localize computational nodes on the deployed grid. With this mechanism, we can determine if two nodes are on the same cluster. This mechanism is used in Grid'BnB to reduce inter-cluster communications. We run experiments on a nationwide grid. With this test bed, we analyze the behavior of a communicant application deployed on a large-scale grid that solves the flow-shop problem.

show abstract

“…Conceptually it is relatively easy to provide parallel implementations of B&B. Many previous work use the master-worker paradigm [4,5].…”

Section: Principlesmentioning

confidence: 99%

“…ParadisEO splits the optimization problem in tasks. Then, the task allocation is handled by MW [4], a tool for scheduling master-worker applications over Condor [14], which is a grid resource manager. Unlike Grid'BnB, ParadisEO just provides mechanisms for searching algorithms.…”

Section: Related Workmentioning

confidence: 99%

Grid’BnB: A Parallel Branch and Bound Framework for Grids

Caromel

Costanzo

Baduel

et al. 2007

High Performance Computing – HiPC 2007

View full text Add to dashboard Cite

show abstract

“…Different inter-task synchronization constraints impact the job performance. We have recently begun exploration of the effectiveness of replication for a Master-Worker workload [15], but for now consider the following workload models:…”

Section: B Parallel Job Modelsmentioning

confidence: 99%

Improving Speedup and Response Times by Replicating Parallel Programs on a SNOW

Ghare

Leutenegger

2005

Job Scheduling Strategies for Parallel Processing

View full text Add to dashboard Cite

Idle computation cycles of a shared network of workstations are increasingly being used to run batch parallel programs. For one common paradigm, the batch program task running on an idle workstation is preempted when the owner reclaims the workstation. This owner interference has a considerable impact on the execution time of a batch program, especially in the case of large parallel programs. Replication of batch program tasks has been used to reduce the impact of owner interference. We show analytically that reclamation can significantly improve parallel program speedup. Perhaps surprisingly, replication can also improve efficiency for certain workloads. We present analysis to quantify the amount of speedup and efficiency improvement. Furthermore, we provide analysis to help determine whether extra available workstations should be used for increasing job parallelism or for task replication.

show abstract

“…For this purpose, we have used Condor as a resource management on a grid-computing environment [4] and the MW library [5] so as to easily implement our master-worker scheme. In the following section, we briefly outline certain basic ideas on Condor and MW.…”

Section: Implementation Issuesmentioning

confidence: 99%

Evolutionary Optimization Techniques on Computational Grids

Abdalhaq

Cortés

Margalef

et al. 2002

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Optimization of complex objective functions such as environmental models is a compute-intensive task, difficult to achieve by classical optimization techniques. Evolutionary techniques such as genetic algorithms present themselves as the best alternative to solving this problem. We present a friendly optimization framework for complex objective function on a computational grid platform, which allows easy incorporation of new optimization strategies. This framework was developed using the MW library and the Condor system. The framework architecture is described, and a case study of a forest-fire propagation simulator is then analyzed.

show abstract

An enabling framework for master-worker applications on the Computational Grid

Cited by 149 publications

References 11 publications

Grid’BnB: A Parallel Branch and Bound Framework for Grids

Grid’BnB: A Parallel Branch and Bound Framework for Grids

Improving Speedup and Response Times by Replicating Parallel Programs on a SNOW

Evolutionary Optimization Techniques on Computational Grids

Contact Info

Product

Resources

About