Runtime Model-Based Privacy Checks of Big Data Cloud Services

Cloud services have to comply with privacy policies when storing or processing data. As cloud services become increasingly data-intensive, e.g., in the case of big data analytics, data privacy concerns become more critical and challenging to address. In particular, data may only be processed at certain geo-locations. However, the actual geo-locations of the many storage and compute nodes involved in big data processing is dynamically selected during runtime. In addition, the execution of concrete data processing tasks may change data classifications from, e.g., personal to anonymized data. Thus, privacy policy checks for big data cloud services have to consider information about the actual nodes and data processing tasks at runtime. The proposed approach R-PRIS monitors cloud services to derive and maintain typed runtime models providing the aforementioned information. R-PRIS checks the typed runtime models against privacy policies by employing a data-classification-aware search. The evaluation of R-PRIS, performed on Amazon Web Services (including Hadoop), indicates that the approach may efficiently and timely detect privacy violations in big data cloud services.