Click here for full text:
The Case for Data Assurance
Arlitt, Martin; Farkas, Keith I.
Keyword(s): Information Technology Computing Environments; monitoring data; data quality
Abstract: Monitoring data, while used historically primarily for performance analysis and debugging, is increasingly playing a more central role in the control and management of IT environments. Yet, monitoring data contains errors, errors which can and do affect the decisions made by the automatic control and management systems. To support the increasing scale and complexity of IT environments, it is imperative that data quality problems be recognized before automated systems act on the data. We believe this need can be met with a data assurance layer, which sits between the data collectors and consumers (human or software), and which is responsible for informing the data consumers about its quality. In this paper, we lay the groundwork for this layer by examining the imperfections that exist in monitoring data, and then identifying some promising directions for further investigation.
Back to Index