ESTOCADA: Towards Scalable Polystore Systems

Big data applications increasingly involve diverse datasets, conforming to different data models. Such datasets are routinely hosted in heterogeneous stores, each capable of handling one or a few data models, and each efficient for some, but not all, kinds of data processing. Systems capable of exploiting disparate data in this fashion are usually termed polystores. A current limitation of polystores is that applications are written taking into account which part of the data is stored in which store and how. This fails to take advantage of ( i ) possible redundancy, when the same data may be accessible (with different performance) from distinct data stores; ( ii ) previous query results (in the style of materialized views), which may be available in the stores. We propose to demonstrate ESTOCADA [4], a novel approach that can be used in a polystore setting to transparently enable each query to benefit from the best combination of stored data and available processing capabilities. The system leverages recent advances in the area of view-based query rewriting under constraints, which we use to describe the various data models and stored data.