Wednesday, 13 December 2017

Session 2: Best Research Paper Candidates

Uncertainties in the Modeling of Self-adaptive Systems: A Taxonomy and an Example of Availability Evaluation

Authors:

Diego Perez-Palacin (Politecnico di Milano)
Raffaela Mirandola (Politecnico di Milano)

Abstract:

The complexity of modern software systems has grown enormously in the past years with users always demanding for new features and better quality of service. Besides, software is often embedded in dynamic contexts, where requirements, environment assumptions, and usage profiles continuously change. As an answer to this need, it has been proposed the usage of self-adaptive systems. Self-adaptation endows a system with the capability to accommodate its execution to different contexts in order to achieve continuous satisfaction of requirements. Often, self-adaptation process also makes use of runtime model evaluations to decide the changes in the system. However, even at runtime, context information that can be managed by the system is not complete or accurate; i.e, it is still subject to some uncertainties. This work motivates the need for the consideration of the concept of uncertainty in the model-based evaluation as a primary actor, classifies the avowed uncertainties of self-adaptive systems, and illustrates examples of how different types of uncertainties are present in the modeling of system characteristics for availability requirement satisfaction.

DOI: 10.1145/2568088.2568095

Full text: PDF

[#][]

On the Limits of Modeling Generational Garbage Collector Performance

Authors:

Peter Libič (Charles University)
Lubomír Bulej (University of Lugano)
Vojtěch Horký (Charles University)
Petr Tůma (Charles University)

Abstract:

Garbage collection is an element of many contemporary software platforms whose performance is determined by complex interactions and is therefore difficult to quantify and model. We investigate the difference between the behavior of a real garbage collector implementation and a simplified model on a selection of workloads, focusing on the accuracy achievable with particular input information (sizes, references, lifetimes). Our work highlights the limits of performance modeling of garbage collection and points out issues of existing evaluation tools that may lead to incorrect experimental conclusions.

DOI: 10.1145/2568088.2568097

Full text: PDF

[#][]