Replicability of Dynamically Provisioned Scientific Experiments

Abstract

The ability to repeat an experiment, known as replicability, is a basic concept of scientific research and also an important aspect in the field of eScience. The principles of Service Oriented Computing (SOC) and Cloud Computing, both based on high runtime dynamicity, are more and more adopted in the eScience domain. Simulation experiments exploiting these principles introduce significant challenges with respect to replicability. Current research activities mainly focus on how to exploit SOC and Cloud for eScience, while the aspect of replicability for such experiments is still an open issue. In this paper we define a general method to identify points of dynamicity in simulation experiments and to handle them in order to enable replicability. We systematically examine different types of service binding strategies, the main source of dynamicity, and derive a method and corresponding architecture to handle this dynamicity with respect to replicability. Our work enables scientists to perform simulation experiments that exploit the dynamicity and flexibility of SOC and Cloud Computing but still are repeatable.