000006608 001__ 6608
000006608 005__ 20250314084056.0
000006608 0247_ $$2DOI$$a10.1016/j.parco.2008.12.012
000006608 0247_ $$2WOS$$aWOS:000272962600004
000006608 037__ $$aPreJuSER-6608
000006608 041__ $$aeng
000006608 082__ $$a004
000006608 084__ $$2WoS$$aComputer Science, Theory & Methods
000006608 1001_ $$0P:(DE-Juel1)VDB62975$$aBecker, D.$$b0$$uFZJ
000006608 245__ $$aScalable timestamp synchronization for event traces of message-passing applications
000006608 260__ $$aAmsterdam [u.a.]$$bNorth-Holland, Elsevier Science$$c2009
000006608 300__ $$a595 - 607
000006608 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article
000006608 3367_ $$2DataCite$$aOutput Types/Journal article
000006608 3367_ $$00$$2EndNote$$aJournal Article
000006608 3367_ $$2BibTeX$$aARTICLE
000006608 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000006608 3367_ $$2DRIVER$$aarticle
000006608 440_0 $$012681$$aParallel Computing$$v35$$x0167-8191$$y12
000006608 500__ $$aRecord converted from VDB: 12.11.2012
000006608 520__ $$aEvent traces are helpful in understanding the performance behavior of message-passing applications since they allow the in-depth analysis of communication and synchronization patterns. However, the absence of synchronized clocks may render the analysis ineffective because inaccurate relative event timings may misrepresent the logical event order and lead to errors when quantifying the impact of certain behaviors. Although linear offset interpolation can restore consistency to some degree, time-dependent drifts and other inaccuracies may still disarrange the original succession of events - especially during longer runs. The controlled logical clock algorithm accounts for such violations in point-to-point communication by shifting message events in time as much as needed while trying to preserve the length of local intervals. In this article, we describe how the controlled logical clock is extended to collective communication to enable the correction of realistic message-passing traces. We present a parallel version of the algorithm scaling to more than thousand processes and evaluate its accuracy by showing that it eliminates inconsistent inter-process timings while preserving the length of local intervals. (C) 2009 Elsevier B.V. All rights reserved.
000006608 536__ $$0G:(DE-Juel1)FUEK411$$2G:(DE-HGF)$$aScientific Computing$$cP41$$x0
000006608 536__ $$0G:(DE-Juel-1)ATMLPP$$aATMLPP - ATML Parallel Performance (ATMLPP)$$cATMLPP$$x1
000006608 588__ $$aDataset connected to Web of Science
000006608 650_7 $$2WoSType$$aJ
000006608 65320 $$2Author$$aPerformance analysis
000006608 65320 $$2Author$$aEvent tracing
000006608 65320 $$2Author$$aClock synchronization
000006608 7001_ $$0P:(DE-HGF)0$$aRabenseifner, R.$$b1
000006608 7001_ $$0P:(DE-Juel1)VDB1927$$aWolf, F.$$b2$$uFZJ
000006608 7001_ $$0P:(DE-HGF)0$$aLinford, J.$$b3
000006608 773__ $$0PERI:(DE-600)1466340-5$$a10.1016/j.parco.2008.12.012$$gVol. 35, p. 595 - 607$$p595 - 607$$q35<595 - 607$$tParallel computing$$v35$$x0167-8191$$y2009
000006608 8567_ $$uhttp://dx.doi.org/10.1016/j.parco.2008.12.012
000006608 909CO $$ooai:juser.fz-juelich.de:6608$$pVDB
000006608 9131_ $$0G:(DE-Juel1)FUEK411$$bSchlüsseltechnologien$$kP41$$lSupercomputing$$vScientific Computing$$x0
000006608 9141_ $$y2009
000006608 915__ $$0StatID:(DE-HGF)0010$$aJCR/ISI refereed
000006608 9201_ $$0I:(DE-Juel1)JSC-20090406$$gJSC$$kJSC$$lJülich Supercomputing Centre$$x0
000006608 9201_ $$0I:(DE-82)080012_20140620$$gJARA$$kJARA-HPC$$lJülich Aachen Research Alliance - High-Performance Computing$$x1
000006608 970__ $$aVDB:(DE-Juel1)114965
000006608 980__ $$aVDB
000006608 980__ $$aConvertedRecord
000006608 980__ $$ajournal
000006608 980__ $$aI:(DE-Juel1)JSC-20090406
000006608 980__ $$aI:(DE-82)080012_20140620
000006608 980__ $$aUNRESTRICTED
000006608 981__ $$aI:(DE-Juel1)VDB1346