Performance Prediction of Distributed Applications Using Block Benchmarking Methods

Cornea, Bogdan; Bourgeois, Julien

doi:10.1109/pdp.2011.52

Cited by 4 publications

(12 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2) Methodology: dPerf methodology has previously been described in [6], therefore, in this article, we briefly go through the most important aspects of our performance prediction approach. The dPerf approach for doing performance prediction for distributed applications running with the P2PDC decentralized environment is presented in the following.…”

Section: Performance Prediction With Dperf -Methodology and Requirmentioning

confidence: 99%

“…This point in the analysis process is responsible for inserting into the studied AST of calls to the PAPI library for obtaining accurate measurement of time duration. Two ways of performing the automatic static analysis are implemented in dPerf, as explained in [6], but in this paper we only employ the simple block benchmarking technique. The use of benchmarking by block makes it possible for dPerf results to be scaled-up while maintaining accuracy in predicting the performance.…”

Section: Performance Prediction With Dperf -Methodology and Requirmentioning

confidence: 99%

“…The dPerf prediction environment is depicted in Fig. 6 and has previously been presented in [6]. We will briefly enumerate several requirements that must be met so that application performance prediction can be done, and then we quickly remind the most important aspects of dPerf.…”

Section: Hierarchical Task Allocationmentioning

confidence: 99%

“…Traditional prediction tools are not particularly well-suited to large scale peer-to-peer HPC applications. For this reason, we have developed the distributed Performance Prediction environment (dPerf), see [6]. dPerf is inspired by previous work [7] which uses static and semi static analysis of distributed applications that communicate using MPI [8].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

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

Cornea¹,

Bourgeois²,

Nguyen³

et al. 2011

2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and PHD Forum

Self Cite

View full text Add to dashboard Cite

International audienceP2PDC is an environment for high performance peer to peer computing that allows direct communication between peers. This environment is based on P2PSAP, a self adaptive communication protocol. P2PDC is suited to the solution of large scale numerical simulation problems via distributed iterative methods. dPerf is a performance prediction environment for parallel and distributed applications, with primary interest in programs written in C, C++, Fortran for P2PDC. The dPerf performance prediction tool makes use of static and dynamic analyses combined with trace-based simulation. In this paper, we present a decentralized version of P2PDC and show how dPerf predicts performance for the P2PDC environment. We present new features of P2PDC aimed at making it more scalable and robust. Through experiments with P2PDC and dPerf, we show how to properly choose a peer to peer computing system which can match the computing power of a cluster

show abstract

Section: Performance Prediction With Dperf -Methodology and Requirmentioning

confidence: 99%

Section: Performance Prediction With Dperf -Methodology and Requirmentioning

confidence: 99%

Section: Hierarchical Task Allocationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

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

Cornea¹,

Bourgeois²,

Nguyen³

et al. 2011

2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and PHD Forum

Self Cite

View full text Add to dashboard Cite

show abstract

“…Whereas JProfiler uses tremendously more sophisticated profiling techniques than its predecessors, it still considers the output of its profiling activity as a mere tree widget to indicate the CPU consumption. It appears that a large part of the research conducted in the field of code profiling focuses on reducing the overhead triggered by the code instrumentation and observation .…”

Section: Introductionmentioning

confidence: 99%

Execution profiling blueprints

et al. 2011

View full text Add to dashboard Cite

SUMMARY Although traditional approaches to code profiling help locate performance bottlenecks, they offer only limited support for removing these bottlenecks. The main reason is the lack of detailed visual runtime information to identify and eliminate computation redundancy. We provide three profiling blueprints that help identify and remove performance bottlenecks. The structural distribution blueprint graphically represents the CPU consumption share for each method and class of an application. The behavioral distribution blueprint depicts the distribution of CPU consumption along method invocations and hints at method candidates for caching optimizations. The behavioral evolution blueprint compares profiles of different versions of a software system and highlights performance‐critical changes in the system. These three blueprints helped us to significantly optimize Mondrian, an open source visualization engine. Our implementation is freely available for the Pharo development environment and has been evaluated in a number of different scenarios. Copyright © 2011 John Wiley & Sons, Ltd.

show abstract