Flexible Rule-Based Programming for Autonomic Computing

The ECAP rule programming language DLRL is currently being developed for architecting autonomic systems by coupling deduction and interaction. Three of the fundamental properties of autonomic systems, namely self-configuration, self-optimization and self-healing, are provided by DLRL: high-level program specification that can be userdefined for self-configuration; program introspection that allows to reactively adapt on-line program behavior for self-optimization, and program interaction that provides communication and coordination with the surrounding environment in order to detect deviations from their expected behavior for self-healing. The DLRL, programming model extends pure Prolog by including the modal actions of dynamic logic in the consequent predicates of conditional forward rules. It combines some the well-known refinements of resolution along with syntactically guided control strategies to represent and enact problem specifications dealing with stateless and state-based descriptions. The main contribution of this work consists in showing the benefits for architecting autonomic systems in a single, uniform and expressive mulit-paradigm programming language for rapidly changing demands of complex problems in distributed settings.