Towards Understanding Adaptation Latency in Self-adaptive Systems

A major advantage of service-based and cloud-based systems is their ability to perform self-adaptation. Through self-adaptation, such systems can automatically react to changes and thus ensure the continued satisfaction of their functional and non-functional requirements. Self-adaptation may take non-negligible time (which we term adaptation latency), and during this period the self-adaptive system may exhibit degraded performance or other negative impact. Hence, it is important to understand how long self-adaptations take and what influences the adaptation latency. However, we are not aware of a systematic study of this question in the literature. This paper is a first step in this direction. We present (i) a model of adaptation latency that breaks it down into four components and (ii) a preliminary survey to analyze information about adaptation latency in the available literature on self-adaptive systems. According to our findings, although some components of the adaptation latency are studied in some publications, the whole adaptation delay is seldom considered.