MONA+ Additional Information

Invited Talk "Monitoring and Adaptation in Open-world Software"

Sam Guinea, Politecnico di Milano

Abstract: The functional and non-functional requirements of today's software have sky-rocketed, and software needs to rise to the challenge and become enablers for scenarios that were previously unimaginable. Ubiquitous, context- and situational-aware applications are common, and distributed systems are the norm. Service technologies have appeared on the landscape, bringing distributed ownership to the picture, meaning designers must now wrap their heads around new and extremely complex landscapes. Software now lives in an open world. No longer can good designers constrain, at design time, the complexity and the number of problems that can arise at run time. On the contrary, the open world is made up of dynamic systems, of components and services that constantly enter or exit the field, or that change their functional and non-functional qualities without notice. Our focus shifts heavily to run-time management, and in particular to monitoring and adaptation. Continues monitoring has been introduced since it has become impossible to solely rely on classic static validation and verification techniques. Monitoring means "keeping an eye" on our system, to figure out, as soon as possible, if something is not going as expected. On the other hand, adaptation tries to change how a system is behaving, with the goal of optimizing quality of service. A system can be adapted as an answer to an anomalous situation, to try to "keep things on track", it can be pre-emptive, if the system makes use of prediction mechanisms, or it can simply be performed to take advantage of changes in the context of execution.

In the talk I will pinpoint the main problems that need to be addressed in open-world software, with a focus on monitoring and adaptation. I will briefly present the most prominent existing approaches, and give an update on my own research agenda. Finally, I will attempt to highlight interesting research issues that remain open, and take a look at what could be our next steps going forward.