About Lotus@Runtime Lotus@Runtime is an extensible tool that uses models@runtime to monitor and verify self-adaptive systems. The tool monitors the execution traces generated by a self-adaptive system and annotates the probabilities of occurrence of each system action on their respective transition on the system model, which is represented by a Probabilistic Labelled Transition System (PLTS). Then, runtime checks of a set of reachability properties are performed against the updated model. If a property is violated, the self-adaptive system can be informed by a notification mechanism provided by Lotus@Runtime.
In the Lotus@Runtime, an adaptation flow, illustrated in Figure below, starts by monitoring the execution trace generated by a self-adaptive system. The traces are captured during the monitoring phase by the MonitorComponent. Then, during the model update process, the LotusModelComponent calculates the probabilities of each transition and updates the model with the new probabilities at runtime. After, the ModelCheckerComponent performs runtime checks of the user defined properties against the updated probabilistic software behavior model. If a property is violated, then NotifierComponent triggers a notification about that violation. Regardless of whether or not a violation occurs, the monitoring process should keep running continuously. To receive the violations that may occur, developers must implement the interface ViolationHandler at the planning phase. Otherwise, the notifications will be lost.
More Information: