A Trace Macroscopic Description based on Time Aggregation

Today, because of computing system complexity, it is required to trace application executions to understand their behavior. Visualization techniques provide some help in representing their content, but their scalability is limited both because of human perception and bounded screen resolution. To solve this issue, we propose a visualization based on time aggregation that provides a concise overview of a trace whatever its size. The level of details in this visualization can be configurable by users who can adjust the compromise between concision (gain from aggregation) and information loss. They can then refine their analysis by zooming in an interesting part and choosing a less aggregated overview for this interesting part. This visualization is implemented in our tool, Ocelotl, which enables users to interact with this visualization by changing the selected time interval and its aggregation settings dynamically. The results presented in this paper show that the technique can help users correctly identify anomalies in very large trace files composed of up to forty million events.

[1]  Barton P. Miller,et al.  The Paradyn Parallel Performance Measurement Tool , 1995, Computer.

[2]  Allen D. Malony,et al.  ParaProf: A Portable, Extensible, and Scalable Tool for Parallel Performance Profile Analysis , 2003, Euro-Par.

[3]  Matthias S. Müller,et al.  The Vampir Performance Analysis Tool-Set , 2008, Parallel Tools Workshop.

[4]  Toni Cortes,et al.  PARAVER: A Tool to Visualize and Analyze Parallel Code , 2007 .

[5]  Arie van Deursen,et al.  Execution trace analysis through massive sequence and circular bundle views , 2008, J. Syst. Softw..

[6]  James M. Wilson,et al.  Gantt charts: A centenary appreciation , 2003, Eur. J. Oper. Res..

[7]  Ben Shneiderman,et al.  The eyes have it: a task by data type taxonomy for information visualizations , 1996, Proceedings 1996 IEEE Symposium on Visual Languages.

[8]  Charles F. Hockett,et al.  A mathematical theory of communication , 1948, MOCO.

[9]  David H. Bailey,et al.  The Nas Parallel Benchmarks , 1991, Int. J. High Perform. Comput. Appl..

[10]  Jean-Marc Vincent,et al.  The Best-Partitions Problem: How to Build Meaningful Aggregations , 2013, 2013 IEEE/WIC/ACM International Joint Conferences on Web Intelligence (WI) and Intelligent Agent Technologies (IAT).

[11]  B. de Oliveira Stein,et al.  Pajé trace file format , 2003 .

[12]  Allen D. Malony,et al.  PerfExplorer: A Performance Data Mining Framework For Large-Scale Parallel Computing , 2005, ACM/IEEE SC 2005 Conference (SC'05).

[13]  Michael T. Heath,et al.  Visualizing the performance of parallel programs , 1991, IEEE Software.

[14]  Jean-Marc Vincent,et al.  Trace Management and Analysis for Embedded Systems , 2013, 2013 IEEE 7th International Symposium on Embedded Multicore Socs.

[15]  R. A. Leibler,et al.  On Information and Sufficiency , 1951 .

[16]  Jesús Labarta,et al.  John von Neumann Institute for Computing Scalability of Visualization and Tracing Tools , 2022 .

[17]  Ewing L. Lusk,et al.  Early Experiments with the OpenMP/MPI Hybrid Programming Model , 2008, IWOMP.

[18]  Jean-Marc Vincent,et al.  Identification of international media events by spatial and temporal aggregation of RSS flows of newspapers , 2013 .

[19]  Jean-Marc Vincent,et al.  Detection and analysis of resource usage anomalies in large distributed systems through multi‐scale visualization , 2012, Concurr. Comput. Pract. Exp..

[20]  Jacques Chassin de Kergommeaux,et al.  Pajé, an interactive visualization tool for tuning multi-threaded parallel applications , 2000, Parallel Comput..