Performance Prediction in a Decentralized Environment for Peer-to-Peer Computing

Cornea, Bogdan; Bourgeois, Julien; Nguyen, The Tung; Baz, Didier El

doi:10.1109/ipdps.2011.321

Cited by 7 publications

(10 citation statements)

References 33 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The real conditions are taken into account by considering compiler optimizations when computing the prediction. The novelty in dPerf features is the scalability of results which extends previous work on dPerf [2,4]. The development of dPerf is a result of a thorough study of the state of art in performance prediction tools which shows that several aspects are hardly addressed by existing methods.…”

Section: Introductionmentioning

confidence: 70%

Scalable Performance Predictions of Distributed Peer-to-Peer Applications

Cornea

Bourgeois

Nguyen

et al. 2012

2012 IEEE 14th International Conference on High Performance Computing and Communication &Amp; 2012 IEEE 9th International Confe

Self Cite

View full text Add to dashboard Cite

International audienceRecently, a new environment for high performance peer-to-peer distributed computing was proposed. This environment , named P2PDC, addresses stable or volatile systems communicating in a decentralized manner using the self-adaptive protocol P2PSAP. P2PDC is devoted to task parallel applications like numerical simulation problems or optimization problems solved via parallel or distributed iterative algorithms. For distributed applications meant to run with P2PDC, a performance prediction tool named dPerf was proposed. dPerf combines static and dynamic analysis with trace-based simulation to provide scientist with information about the execution of their large scale numerical simulation applications. dPerf addresses real parallel and distributed numerical simulation and optimisation applications written in C, C++ or Fortran for P2PDC. This paper introduces an enhancement of the dPerf tool which provides scalable performance prediction results. Scaling is done with respect to (i) network configuration and (ii) number of peers. Scaling predictions based on network configuration is achieved through trace-based simulation, where various architectures can be studied. Scaling predictions based on the number of peers implies analyzing the communication topology and modifying trace files prior to simulation. We present experimental results obtained for the obstacle problem, a C/P2PDC implementation of the code used in mechanics and finance. Prediction for this application is computed under real conditions, with a reduced slowdown and by providing user with scalable results

show abstract

Section: Introductionmentioning

confidence: 70%

Scalable Performance Predictions of Distributed Peer-to-Peer Applications

Cornea

Bourgeois

Nguyen

et al. 2012

2012 IEEE 14th International Conference on High Performance Computing and Communication &Amp; 2012 IEEE 9th International Confe

Self Cite

View full text Add to dashboard Cite

show abstract

“…The P2PSAP protocol can also reconfigure itself dynamically to match the very nature of a given algorithm like in the case of the distributed solution of nonstationary problems whereby some synchronizations are needed at every time steps even if an asynchronous scheme of computation has been selected to solve each subproblem. Reference is made to [11] and [6] for studies related to the centralized and decentralized versions of the P2PDC environment, respectively.…”

Section: Related Workmentioning

confidence: 99%

“…The P2PSAP self-adaptive communication protocol [5] and P2PDC distributed computing environment [6] have originally been designed for peer-to-peer HPC applications. The P2PSAP self-adaptive communication protocol is based on the Cactus framework whereby composite protocols and micro-protocols are combined in order to build a desired communication protocol [7].…”

Section: Introductionmentioning

confidence: 99%

Distributed Iterative Solution of Numerical Simulation Problems on Infiniband and Ethernet Clusters via the P2PSAP Self-Adaptive Protocol

Tembo

Nguyen

Baz

2013

2013 21st Euromicro International Conference on Parallel, Distributed, and Network-Based Processing

View full text Add to dashboard Cite

Abstract-The distributed iterative solution of numerical simulation problems on Infiniband or Ethernet Clusters via the P2PDC environment is studied. The P2PDC decentralized environment is dedicated to task parallel applications. It has been designed for the solution of large scale numerical simulation problems via distributed iterative algorithms. The P2PDC environment is based on the P2PSAP self-adaptive communication protocol. New functionalities of the P2PSAP communication protocol aimed at using Infiniband clusters are presented. A series of computational results is presented and analyzed.

show abstract

“…We note that these steps can be iterated several times in order to improve prediction. Reference is made to [12] to [14] for previous work on performance prediction of HPC applications on distributed computing infrastructures.…”

Section: B the Hpc Infrastructure 1) The Brokermentioning

confidence: 99%