Replicated condition monitoring

Abstract

A condition monitoring system tracks real-world variables and alerts users when a predefined condition becomes true, e.g., when stock price drops, or when a nuclear reactor overheats. Replication of monitoring servers can reduce the probability that an important alert is missed. However, replicated independent servers can sometimes report ``conflicting'' alerts to the user, causing confusion. In this paper, we study the problem of replicated condition monitoring. We identify and formally define three desirable properties of a replicated system, namely, orderedness, consistency, and completeness. We propose new monitoring algorithms that enforce some or all of the desired properties in different scenarios.