Parallel query processing in a polystore

The blooming of different data stores has made polystores a major topic in the cloud and big landscape. As amount grows rapidly, it becomes critical to exploit inherent parallel processing capabilities underlying platforms. To fully achieve this, polystore should: (i) preserve expressivity each store’s native query or scripting language (ii) leverage distributed architecture enable integration, i.e. joins, on top retrieval partitioned datasets. In this paper, we address these points by: using polyglot approach CloudMdsQL that allows queries be expressed as inline scripts combined with SQL statements for ad-hoc integration incorporating within LeanXcale engine, thus allowing processed at store shards. addition, (iii) efficient optimization techniques, such bind join, can take place improve performance selective joins. We evaluate benefits exploiting parallelism combination high through our experimental validation.