SQL-On-Hadoop Systems: Evaluting Performance of Polybase for Big Data Processing

Performance of the bridge PolyBase SQL Server for working with hybrid data stores has been investigated. The analysis of the existing bridges for joint processing of SQL-and NoSQL data with using of a Hadoop cluster has been carried out. Technologies for query optimization when using PolyBase have been represented. The paradigm of data splitting and applying the pushdown procedure has been considered. Results of using the procedure predicate pushdown for the suggested metrics of PolyBase-Hadoop system performance have been considered for examples of processing test data of large volumes.