Statistical analysis with missing data
Statistical analysis with missing data
Communications of the ACM - Supporting community and building social capital
Time Series Analysis, Forecasting and Control
Time Series Analysis, Forecasting and Control
HP-UX 11i Tuning and Performance (2nd Edition)
HP-UX 11i Tuning and Performance (2nd Edition)
End of software, the: finding security, flexibility, and profit in the on demand future
End of software, the: finding security, flexibility, and profit in the on demand future
Fa: A System for Automating Failure Diagnosis
ICDE '09 Proceedings of the 2009 IEEE International Conference on Data Engineering
Do you know your IQ?: a research agenda for information quality in systems
ACM SIGMETRICS Performance Evaluation Review
Mining console logs for large-scale system problem detection
SysML'08 Proceedings of the Third conference on Tackling computer systems problems with machine learning techniques
| Hi-index | 0.00 |
Efforts to reduce the cost of ownership for enterprise IT environments are spurring the development and deployment of data-driven management tools. Yet, IT data is imperfect and these imperfections can lead to inappropriate decisions that have significant technical and business consequences. In this paper, we begin by raising awareness of this problem through examples of the imperfections that occur, and a discussion of their causes and implications on IT management tasks. We then introduce a systematic approach for addressing such imperfections. Our approach allows best practices to be readily shared, simplifies the construction of IT data assurance solutions, and allows context-specific corrections to be applied until the root cause(s) of the imperfections can be fixed. To demonstrate the value of our solution, we describe a capacity planning use case. Application of our solution to an ongoing capacity planning effort reduced the (human) planner's time requirements by ≈3x to ≈6 hours, while enabling him to evaluate the data quality of ≈5x more applications and for 9 imperfection types rather than 1.