Performance profiling generates measurement overhead during parallel program
execution. Measurement overhead, in turn, introduces intrusion in a program's
runtime performance behavior. Intrusion can be mitigated by controlling
instrumentation degree, allowing a tradeoff of accuracy for detail. Alternatively,
the accuracy in profile results can be improved by reducing the intrusion error due
to measurement overhead. Models for compensation of measurement overhead in parallel
performance profiling are described. An approach based on rational reconstruction is
used to understand properties of compensation solutions for different parallel
scenarios. From this analysis, a general algorithm for on-the-fly overhead
assessment and compensation is derived.
Bailey, D.
,
Harris, T.
,
Saphir, W.
,
van der Wijngaart, R.
,
Woo, A.
, and
Yarrow, M.
(1995). The NAS parallel benchmarks 2.0, Technical Report
NAS-95-020,
NASA Ames Research Center
.
2.
Bell Laboratories.
(1979). prof command, in Unix Programmer's Manual, Section
1,
Murray Hill, NJ
.
3.
Bronevetsky, G.
,
Marques, D.
,
Pingali, K.
, and
Stodghill, P.
(2003a). Automated application-level checkpointing of mpi
programs, in Principles and Practice of Parallel Programming (PPoPP).
4.
Bronevetsky, G.
,
Marques, D.
,
Pingali, K.
, and
Stodghill, P.
(2003b). Collective operations in an application-level fault
tolerant mpi system, in International Conference on
Supercomputing (
ICS
).
5.
Browne, S.
,
Dongarra, J.
,
Garner, N.
,
Ho, G.
, and
Mucci, P.
(2000).
A portable programming interface for performance evaluation on modern processors
, International Journal of High Performance Computing
Applications, 14(3):
189— 204
.
6.
Fahringer, T.
and
Seragiotto, C.
(2002).
Experience with aksum: A semi-automatic multi-experiment performance
analysis tool for parallel and distributed applications
, in Workshop on Performance Analysis and Distributed
Computing .
7.
Graham, S.
,
Kessler, P.
, and
McKusick, M.
(1982).
gprof: A call graph execution profiler
, in SIGPLAN Symposium on Compiler Construction, pp. 120—126.
8.
Ibm.
(n.d.). Profiling parallel programs with xprofiler, in IBM
parallel environment for AIX:
Operation and use
, vol. 2.
9.
Janssen, C.
(n.d.). The visual profiler, URLhttp://aros.ca.sandia.gov/cljanss/perf/vprof/.
10.
Knuth, D.
(1971).
An empirical study of fortran programs
, Software Practice and Experience, 1:
105—133
.
11.
Kranzlmüller, D.
(2000). Event graph analysis for debugging massively parallel
programs, Ph.D. thesis, Johanne Kepler Universität
Linz
.
12.
Malony, A.
(1991a). Event based performance perturbation: A case
study, in Principles and Practices of Parallel Programming (PPoPP),
pp. 201—212.
13.
Malony, A.
(1991b). Performance observability, Ph.D. thesis,
University of Illinois at Urbana-Champaign
.
14.
Malony, A.
and
Reed, D.
(1991).
Models for performance perturbation analysis
, in ACM/ONR Workshop on Parallel and Distributed
Debugging , pp. 1—12.
15.
Malony, A.
,
Reed, D.
, and
Wijshoff, H.
(1992).
Performance measurement intrusion and perturbation analysis
, IEEE Transactions on Parallel and Distributed Systems ,
3(4):
433—450
.
16.
Malony, A.
and
Shende, S.
(2004).
Overhead compensation in performance profiling
, in Euro-Par Conference, Lecture Notes in Computer
Science, vol. 3149, pp. 119—132, Springer.
17.
Malony, A.
and
Shende, S.
(2005).
Models for on-the-fly compensation of measurement overhead in parallel
performance profiling
, in Euro-Par Conference, Lecture Notes in Computer
Science, vol. 3648, pp. 72—82, Springer.
18.
Mellor-Crummey, J.
,
Fowler, R.
, and
Marin, G.
(2002).
Hpcview: A tool for top-down analysis of node performance
, Journal of Supercomputing, 23:
81—104
.
19.
Mucci, P.
(n.d.) Dynaprof, URLhttp://www.cs.utk.edu/mucci/dynaprof.
20.
Reed, D.
,
Rose, L.D.
, and
Zhang, Y.
(1998). Svpablo: A multi-language performance analysis
system, in International Conference on Performance Tools, pp. 352—355.
21.
Rose, L.D.
(2001).
The hardware performance monitor toolkit
, in Euro-Par Conference, Lecture Notes in Computer
Science, vol. 2150, pp. 122—131, Springer.
22.
Sarukkai, S.
and
Malony, A.
(1993). Perturbation analysis of high-level instrumentation
for spmd programs, in Principles and Practices of Parallel
Programming (PPoPP), pp. 44—53.
23.
Shende, S.
and
Malony, A.
(2006).
The tau parallel performance system
, International Journal of High Performance Computing
Applications, 20(2):
287—331
.
24.
Vetter, J.
(2002).
Dynamic statistical profiling of communication activity in distributed applications
, in
ACM SIGMETRICS Joint International Conference on Measurement and Modeling
of ComputerSystems.
25.
Williams, W.
,
Hoel, T.
, and
Pase, D.
(1994). The mpp apprentice performance tool: Delivering the
performance of the cray t3d, in Programming Environments for
Massively Parallel Distributed Systems,
North-Holland
.
