A mathematical model for assessing configuration drift in it systems

This paper introduces a mathematical model for quantitatively assessing configuration drift in IT infrastructure. Each configuration check is represented by a normalized compliance scale, supporting both binary and graded parameters. Checks are grouped into non-overlapping categories to ensure interpretability and enable decomposition of results. The model defines a hierarchical system of weights that allows aggregation from individual checks to nodes, roles, and clusters. Formal properties of the compliance index and the drift index are established, including boundedness, additivity, and robustness to scale transformations. Guidelines are provided for designing atomic checks and consistent measurement scales, and a demonstration example illustrates the calculation of the drift index for a server with multiple parameter categories. The proposed approach provides a rigorous theoretical foundation for further development of monitoring and management systems for configuration drift.

Authors: K. O. Moiseev, A. O. Trofimova

Direction: Informatics, Computer Technologies And Control

Keywords: configuration drift, IT infrastructure, normalized metric, compliance index, aggregation, hierarchical weights, measurement scales, math model

---

View full article