Towards optimal risk-aware security compliance of a large IT system

Abstract

A modern information technology (IT) system may consist of thousands of servers, software components and other devices. Operational security of such a system is usually measured by the compliance of the system with a group of security policies. However, there is no generally accepted method of assessing the risk-aware compliance of an IT system with a given set of security policies. The current practice is to state the fraction of non-compliant systems, regardless of the varying levels of risk associated with violations of the policies and their exposure time windows. We propose a new metric that takes into account the risk of non-compliance, along with the number and duration of violations. This metric affords a risk-aware compliance posture in a single number. It is used to determine a course of remediation, returning the system to an acceptable level of risk while minimizing the cost of remediation and observing the physical constraints on the system, and the limited human labor available. This metric may also be used in the course of the normal operation of the IT system, alerting the operators to potential security breaches in a timely manner. © 2013 Springer-Verlag.