Managing big data using specialized system polar

Abstract

The research is dedicated to the usage of specialized systems for the big data analytics. In the research, the authors focus on the problem of a big data analytics: the fast search, reading and editing the data stored in systems operating with the data containing a great number (ex., about a billion) of records. The methods suggested in the current report have large opportunities of practical applications. The report extends the use of specialized databases. In particular, research was devoted to the use of Object Relational Mapping (ORM) technology for not relational databases. All operations were performed on specialized Polar system, the original implementation of Institute of Information Systems of Siberian Branch of the Russian Academy of Science. The authors have implemented complex ORM Polar system and performed the research to increase its efficiency. The suggested method of the disk space optimization provides significant accelerating of the data search for not-indexed fields. The original scheme of data placing in the Polar system for ORM solution considerably increases the overall performance of the generalized ORM Polar system. All received results were compared to the similar results received for some popular relational DBMS. The executed research confirmed prospects of Polar system use as DBMS for big data, and also proved the possibility of the effective use of ORM technology for it.