Comparing causes of system failure
Microprocessing and Microprogramming
Computer event monitoring and analysis
Computer event monitoring and analysis
Recognition of error symptoms in large systems
ACM '86 Proceedings of 1986 ACM Fall joint computer conference
Dependability Measurement and Modeling of a Multicomputer System
IEEE Transactions on Computers
Design and evaluation of an on-line predictive diagnostic system
Design and evaluation of an on-line predictive diagnostic system
Derivation and Calibration of a Transient Error Reliability Model
IEEE Transactions on Computers
A comparative analysis of event tupling schemes
FTCS '96 Proceedings of the The Twenty-Sixth Annual International Symposium on Fault-Tolerant Computing (FTCS '96)
Critical event prediction for proactive management in large-scale computer clusters
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Performance Implications of Periodic Checkpointing on Large-Scale Cluster Systems
IPDPS '05 Proceedings of the 19th IEEE International Parallel and Distributed Processing Symposium (IPDPS'05) - Workshop 18 - Volume 19
Proceedings of the 22nd Conference of the Computer-Human Interaction Special Interest Group of Australia on Computer-Human Interaction
Performance implications of failures in large-scale cluster scheduling
JSSPP'04 Proceedings of the 10th international conference on Job Scheduling Strategies for Parallel Processing
| Hi-index | 0.00 |
Abstract: Event logs can be used effectively to improve computer system availability. Uses include retrospective and predictive diagnosis; fault management; failure rate estimation; and trend analysis. Unfortunately, much of the research to date has been hampered by the lack of suitable event data, and occasionally by the incorrect interpretation of the available data. This research uses one of the largest sets of data, and the most intensive investigation of the monitoring process conducted to date, to examine event monitoring and analysis. 2.35 million events from 193 VAX/VMS systems covering 335 machine years were used. Examples are presented which show that monitoring deficiencies complicate the analyses, consume additional time, and make incorrect conclusions more likely. For example, incorrect handling of bogus timestamps changes the mean time between groups of events by an order of magnitude. An analysis procedure to identify defects is provided, along with design rules to create better quality logs.