Optimal Evidence Collection for Accountability in the Cloud

In multi-tenant cloud services, accountability can be used to strengthen the trust of tenants in the cloud. It provides the reassurance that the processing of personal data hosted in the cloud is done according to tenants' requirements (a.k.a. Accountability obligations). Ensuring accountability requires multiple measures ranging from preventive controls to violation detection and analysis, based on evidences so as to prove that a violation has occurred or to ensure violation judgment. In a complex cloud environment with multi-tenant services, judging violations encounters difficulties due to the plethora of evidences to be analyzed, which may burden the post-violation analysis in terms of latency and workloads. In this work, we offer a method ensuring the collection of the necessary and minimal (optimal) evidences and avoiding re-evaluating all of them for each violated obligation. Basically, we use a linear program-with an objective function under a set of constraints-and we perform actions in order to obtain optimal evidences elements for the judgment. Finally, our approach is implemented and the results of our experiments highlight its feasibility.