Trace-based parallel performance overhead compensation

Authors:
Felix Wolf;Allen D. Malony;Sameer Shende;Alan Morris
Affiliations:
Innovative Computing Laboratory, University of Tennessee;Department of Computer and Information Science, University of Oregon;Department of Computer and Information Science, University of Oregon;Department of Computer and Information Science, University of Oregon
Venue:
HPCC'05 Proceedings of the First international conference on High Performance Computing and Communications
Year:
2005

Citing 9
Cited 4

Event-based performance perturbation: a case study

PPOPP '91 Proceedings of the third ACM SIGPLAN symposium on Principles and practice of parallel programming
Perturbation analysis of high level instrumentation for SPMD programs

PPOPP '93 Proceedings of the fourth ACM SIGPLAN symposium on Principles and practice of parallel programming
Time, clocks, and the ordering of events in a distributed system

Communications of the ACM
An Adaptive Cost System for Parallel Program Instrumentation

Euro-Par '96 Proceedings of the Second International Euro-Par Conference on Parallel Processing - Volume I
Monitor Overhead Measurement of MPI Applications with SKaMPI

Proceedings of the 6th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Systematic Assessment of the Overhead of Tracing Parallel Programs

PDP '96 Proceedings of the 4th Euromicro Workshop on Parallel and Distributed Processing (PDP '96)
Performance observability

Performance observability
Automatic performance analysis of hybrid MPI/OpenMP applications

Journal of Systems Architecture: the EUROMICRO Journal - Special issue: Evolutions in parallel distributed and network-based processing
An Algebra for Cross-Experiment Performance Analysis

ICPP '04 Proceedings of the 2004 International Conference on Parallel Processing

The ghost in the machine: observing the effects of kernel operation on parallel application performance

Proceedings of the 2007 ACM/IEEE conference on Supercomputing
Interconnection network simulation using traces of MPI applications

International Journal of Parallel Programming
Trace profiling: Scalable event tracing on high-end parallel systems

Parallel Computing
Towards scalable event tracing for high end systems

HPCC'07 Proceedings of the Third international conference on High Performance Computing and Communications

Quantified Score

Hi-index	0.00

Visualization

Abstract

Tracing parallel programs to observe their performance introduces intrusion as the result of trace measurement overhead. If post-mortem trace analysis does not compensate for the overhead, the intrusion will lead to errors in the performance results. We show that measurement overhead can be accounted for during trace analysis and intrusion modeled and removed. Algorithms developed in our earlier work [5] are reimplemented in a more robust and modern tool, kojak [12] , allowing them to be applied in large-scale parallel programs. The ability to reduce trace measurement error is demonstrated for a Monte-Carlo simulation based on a master/worker scheme. As an additional result, we visualize how local perturbation propagates across process boundaries and alters the behavioral characteristics of non-local processes.