Defining and controlling the heterogeneity of a cluster: The Wrekavoc tool

Authors:
Louis-Claude Canon;Olivier Dubuisson;Jens Gustedt;Emmanuel Jeannot
Affiliations:
Nancy Université, Nancy, France and AlGorille Team, INRIA Nancy - Grand Est, France and Runtime Team, INRIA Bordeaux - Sud-Ouest, France;AlGorille Team, INRIA Nancy - Grand Est, France and Felix Informatique, Laxou, France;AlGorille Team, INRIA Nancy - Grand Est, France;AlGorille Team, INRIA Nancy - Grand Est, France and Runtime Team, INRIA Bordeaux - Sud-Ouest, France
Venue:
Journal of Systems and Software
Year:
2010

Citing 13
Cited 2

Direct bulk-synchronous parallel algorithms

Journal of Parallel and Distributed Computing
Scheduling multithreaded computations by work stealing

Journal of the ACM (JACM)
Chord: A scalable peer-to-peer lookup service for internet applications

Proceedings of the 2001 conference on Applications, technologies, architectures, and protocols for computer communications
Matrix Multiplication on Heterogeneous Platforms

IEEE Transactions on Parallel and Distributed Systems
Peer-to-Peer: Harnessing the Power of Disruptive Technologies

Peer-to-Peer: Harnessing the Power of Disruptive Technologies
XtremWeb: A Generic Global Computing System

CCGRID '01 Proceedings of the 1st International Symposium on Cluster Computing and the Grid
Overview of a Performance Evaluation System for Global Computing Scheduling Algorithms

HPDC '99 Proceedings of the 8th IEEE International Symposium on High Performance Distributed Computing
Simulation of Dynamic Data Replication Strategies in Data Grids

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
Data Replication Strategies in Grid Environments

ICA3PP '02 Proceedings of the Fifth International Conference on Algorithms and Architectures for Parallel Processing
The MicroGrid: Using Online Simulation to Predict Application Performance in Diverse Grid Network Environments

CLADE '04 Proceedings of the 2nd International Workshop on Challenges of Large Applications in Distributed Environments
BOINC: A System for Public-Resource Computing and Storage

GRID '04 Proceedings of the 5th IEEE/ACM International Workshop on Grid Computing
Data Partitioning with a Realistic Performance Model of Networks of Heterogeneous Computers with Task Size Limits

ISPDC '04 Proceedings of the Third International Symposium on Parallel and Distributed Computing/Third International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Networks
SimGrid: A Generic Framework for Large-Scale Distributed Experiments

UKSIM '08 Proceedings of the Tenth International Conference on Computer Modeling and Simulation

Accurate emulation of CPU performance

Euro-Par 2010 Proceedings of the 2010 conference on Parallel processing
The impact of heterogeneous multi-core clusters on graph partitioning: an empirical study

Cluster Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The experimental validation and the testing of solutions that are designed for heterogeneous environments are challenging. We introduce Wrekavoc as an accurate tool for this purpose: it runs unmodified applications on emulated multi-site heterogeneous platforms. Its principal technique consists in downgrading the performance of the platform characteristics in a prescribed way. The platform characteristics include the compute nodes themselves (CPU and memory) and the interconnection network for which a controlled overlay network above the homogeneous cluster is built. In this article we describe the tool, its performance, its accuracy and its scalability. Results show that Wrekavoc is a very versatile tool that is useful to perform high-quality experiments (in terms of reproducibility, realism, control, etc.).