Adaptation and Distribution of Pipeline-Based Context-Aware Web Architectures

Authors:
Michael Hinz;Stefan Pietschmann;Matthias Umbach;Klaus Meissner
Affiliations:
Dresden University of Technology, Germany;Dresden University of Technology, Germany;Dresden University of Technology, Germany;Dresden University of Technology, Germany
Venue:
WICSA '07 Proceedings of the Sixth Working IEEE/IFIP Conference on Software Architecture
Year:
2007

Citing 0
Cited 1

Self-tuning of software systems through dynamic quality tradeoff and value-based feedback control loop

Journal of Systems and Software

Quantified Score

Hi-index	0.00

Visualization

Abstract

The dynamic generation of context-aware applications for the ubiquitous Web is very time consuming and causes significant server load. Therefore, optimizing the efficiency of the publishing process and reducing server load by dynamically configuring system components are key factors for the commercial success of context-aware Web applications. To meet this challenge, this paper presents a framework that dynamically adjusts the architecture of pipeline-based context-aware Web systems either by adapting or omitting server tasks, distributing load within server clusters, or migrating load to client devices. The system adaptation and distribution are performed according to varying load rates and system context states that are monitored permanently. Furthermore, the paper also presents experiences gained by the implementation of prototypes and illustrates several useful application scenarios that are supported by the proposed mechanisms.