salle A. Turing CE4
12 September 2014 - 14h00
A Spatiotemporal Data Aggregation Technique for the Macroscopic Analysis of Large-scale Systems
by Jean-Marc Vincent from LIG
Résumé : Analysts commonly use execution traces collected at runtime to understand the behavior of an application running on distributed and parallel systems. These traces are inspected post mortem using various visualization techniques that, however, do not scale properly for a large number of events. This issue, mainly due to human perception limitations, is also the result of bounded screen resolutions preventing the proper drawing of many graphical objects.
This presentation proposes a visualization technique overcoming such limitations by providing a concise overview of the trace behavior as the result of a spatiotemporal data aggregation process. The experimental results show that this approach can help the quick and accurate detection of anomalies in traces containing up to two hundred million events.
This work comes from strong scientific collaborations with Damien Dosimont (LIG-Moais), Robin Lamarche-Perrin (LIG-Magma, Max Planck Institute), Lucas Mello Schnorr (UFRGS) , Guillaume Huard (LIG-Moais), Claude Grasland (Géographie-Cités/CIST) and was supported by ANR SONGS, ANR Geomedia and Licia laboratory.