Towards a run-time model for data protection in the cloud

Abstract

The protection of sensitive data in the cloud is a challenge of increasing importance. It is made particularly difficult by the complex and dynamic interactions of many entities (hardware and software, as well as organizations and individuals). A model-based approach can be used to reason about these interactions and their impact on data protection during deployment and at run time. The basis for such an approach is a model of all relevant socio-technical cloud entities, which is created during deployment and kept alive at run-time to support adaptations. In this paper, we focus on the meta-model of this model. The meta-model is created during design and instantiated during deployment. We discuss what entities must be present in the meta-model to allow reasoning about data protection. In particular, we discuss to what extent the results of previous cloud modeling efforts can be reused and what extensions are necessary because of the particular requirements of data protection.