Monitoring and visualizing adaptation of autonomic systems at runtime

Monitoring an autonomic system at runtime, which typically contains a large number of nodes operating in highly dynamic and open-ended environments, is very challenging for software architects. Solid software engineering methods and tools to support this process are therefore highly required. This paper proposes a novel monitoring and visualization framework for autonomic systems at runtime. Our approach is illustrated by an Eclipse plug-in for tracing the runtime awareness and adaptation capabilities using graph-like representation. A key benefit here is to provide feedback to the engineer about the behavior of the complex awareness mechanism used, thus helping the system evaluation process. We validate and assess our approach and plug-in with a concrete application scenario from swarm robotics domain.