Microelectronics Reliability, Vol. 51, Issue. 2, PP. 263-269, Feb, 2011

Automatic data mining for telemetry database of computer systems

Michael Pecht1 , C.H. Wu2, C.H. Yang2, S.C. Lo3, N. Vichare4, E. Rhem4
1CALCE, Center for Advanced Life Cycle Engineering, Department of Mechanical Engineering, University of Maryland, College Park, Maryland 20740, USA
2Graduate Institute of Mechanical and Electrical Engineering, National Taipei University of Technology, Taipei, Taiwan
3Graduate Institute of Information and Logistics Management, National Taipei University of Technology, Taipei, Taiwan
4Dell Client Systems Reliability Engineering, Dell Computer Corp., Austin, TX, USA


This study focuses on the development of automation platform for performing data mining on a telemetry database for computer systems. It is common for computer systems to encounter failures in an unexpected manner. It is therefore valuable to have prognostics capability for computer systems to minimize the effects of unexpected system failure. Data acquisition schemes employing telemetry techniques are considered the most effective method for collection of in-service information for computer systems. Analysis of an enormous telemetry database of high complexity must be completed before useful knowledge can be extracted. In this research, an automatic data mining platform is reported for the extraction of useful knowledge from the telemetry database. This paper describes the structure and basic theories underlying the data mining of the telemetry database. Also, an automatic computer program capable of performing database management, filtering, data analysis, and reporting is described. Some useful data generated by the platform are reported for the telemetry database.

This article is available online here and to CALCE Consortium Members for personal review.

[Home Page] [Articles Page]
Copyright © 2010 by CALCE and the University of Maryland, All Rights Reserved