Learning and evolution in dynamic software product lines

A Dynamic Software Product Line (DSPL) aims at managing run-time adaptations of a software system. It is built on the assumption that context changes that require these adaptations at run-time can be anticipated at design-time. Therefore, the set of adaptation rules and the space of configurations in a DSPL are predefined and fixed at design-time. Yet, for large-scale and highly distributed systems, anticipating all relevant context changes during design-time is often not possible due to the uncertainty of how the context may change. Such design-time uncertainty therefore may mean that a DSPL lacks adaptation rules or configurations to properly reconfigure itself at run-time. We propose an adaptive system model to cope with design-time uncertainty in DSPLs. This model combines learning of adaptation rules with evolution of the DSPL configuration space. It takes particular account of the mutual dependencies between evolution and learning, such as using feedback from unsuccessful learning to trigger evolution. We describe concrete steps for learning and evolution to show how such feedback can be exploited. We illustrate the use of such a model with a running example from the cloud computing domain.