An approach to cluster system task flow monitoring, analysis, and visualization

Authors

  • A.V. Adinetz
  • P.A. Bryzgalov
  • Vad.V. Voevodin
  • S.A. Zhumatiy
  • D.A. Nikitenko

Keywords:

parallel computing
monitoring
tasks flow
computing cluster

Abstract

Big cluster systems are spreading wide, so that the efficiency of use of such systems is a very actual task for now. In order to solve this task, it is needed to identify the efficiency problems appearing during the task execution, to notify users about appeared problems, and to suggest possible ways to resolve them. This can be achieved by the continuous monitoring of running tasks and by data analysis. This paper discusses an approach to solve these tasks and describes a working prototype.

New computational technologies

Downloads

Published

2011-11-10

Issue

Vol. 12 (2011): Issue 4.

Section

Section 2. Programming

Author Biographies

A.V. Adinetz

Lomonosov Moscow State University,
Research Computing Center
• Junior Researcher

P.A. Bryzgalov

Lomonosov Moscow State University,
Research Computing Center
• Researcher

Vad.V. Voevodin

Lomonosov Moscow State University,
Research Computing Center
• Junior Researcher

S.A. Zhumatiy

Lomonosov Moscow State University,
Research Computing Center
• Senior Researcher

D.A. Nikitenko

Lomonosov Moscow State University,
Research Computing Center
• Junior Researcher

References

  1. Hewitt E. Cassandra: the definitive guide. Cambridge: O’Reilly Media, 2010.
  2. White T. Hadoop: the definitive guide. Cambridge: O’Reilly Media, 2009.
  3. Pig project (http://pig.apache.org/).