Jahresbericht 2021

Abstract

Architecture-based self-adaptive systems abstract the observed behavior of the running system into features of an architectural model, this makes it possible for the adaptation engine to reason about the changes that should be made to a system using variety of existing architectural analysis techniques. There are various ways how self-adaptation following the MAPE-K feedback loop and in particular the analyzing and planning phases of the loop can be realized. Rule-based approaches prescribe the adaptation to be executed if the system or environment satisfy certain conditions and result in scalable solutions, however, with often only satisfying adaptation decisions. In con-trast, utility-driven approaches determine optimal adaptation decisions by using an often costly optimization step, which typically does not scale well for larger problems. We propose a rule-based and utility-driven approach that achieves the beneficial properties of each of these directions such that the adaptation decisions are optimal while the computation remains scalable as an expensive optimization step can be avoided. The approach can be used for the architecture-based self-healing of large software systems. In our approach, we model the dynamic architecture of the self-adaptive system as a graph. Natural state of the system as well as the abstract syntax of the runtime models of the software are depicted via an annotated graph. We apply architectural utility functions in which any possible architectural configuration of the system is mapped to a scalar value.