Performing astronomical data analysis using only personal computers is becoming impractical for the very large data sets produced nowadays. As analysis is not a task that can be automatized to its full extent, the idea of moving processing where the data is located means also moving the whole scientific process towards the archives and data centers. Using Jupyter Notebooks as a remote service is a recent trend in data analysis that aims to deal with this problem, but harnessing the infrastructure to serve the astronomer without increasing the complexity of the service is a challenge. In this paper we present the architecture and features of JOVIAL, a Cloud service where astronomers can safely use Jupyter notebooks over a personal space designed for high-performance processing under the high-availability principle. We show that features existing only in specific packages can be adapted to run in the notebooks, and that algorithms can be adapted to run across the data center without necessarily redesigning them.