Detection and reaction to unplanned operational events in large scale distributed real-time embedded systems

Mission-critical distributed real-time embedded (DRE) systems are becoming quality of service (QoS) enabled, providing predictable, adaptable, and managed behaviors. Due to the dynamic and distributed nature of DRE systems, unplanned events can occur that could significantly impact the behavior and performance of these systems. In order to effectively manage unplanned events in DRE systems, we first clearly define what unplanned events are and how they differ from planned events. A planned event often has known signatures that we can probe for or monitor and is expected to happen during the operation of the system, even if it is "abnormal" behavior. Our discussion on QoS management of unplanned events primarily focuses on unpredictable events rather than unexpected events. The QuO middleware framework provides a template implementation base for designing and instantiating adaptation strategies for both planned and unplanned events independent of functional code, and the resulting QoS management code can be intelligently weaved into the functional code for the system and its applications.