E-GENMR: Enhanced Generalized Query Processing using Double Hashing Technique through MapReduce in Cloud Database Management System

Abstract

Big Data, Cloud computing and Data Science is the booming future of IT industries. The common thing among all the new techniques is that they deal with not just Data but Big Data. Users store various kinds of data on cloud repositories. Cloud Database Management System deals with these large sets of data. Cloud Database service provider deals with many obstacles while providing various service. Amongst all the challenges processing of large amount of data, interoperability and security are the major concerns that are explained in this study. Enhanced Generalized Query Processing through MapReduce (E-GENMR) is a prototype model that provides solution for these problems. Firstly, traditional approaches are not suitable for processing such gigantic amount of data as they are not able to handle such amount of data. Various solutions have been developed such as Hadoop, MapReduce Programming codes, HIVE, PIG etc. but these technologies don't provide solution for these problems at the same time and moreover users are not compatible with these latest technologies like MapReduce codes. E-GENMR provides interoperability as it takes queries written in various RDBMS forms like SQL Server, ORACLE, DB2, MYSQL and convert into MapReduce codes as they are considered to be the efficient way for processing large data. Secondly, Client's data is stored in encrypted form and processing is done on this data hence it ensures the security aspect. Indexing plays a very important role in processing queries, in E-GENMR indexing is implemented using closed double hashing technique. We compared various query processing time of E-GENMR for encrypted data and unencrypted data. A comparison of various queries has been done to evaluate the performance of E-GENMR with latest techniques like Hadoopdb, SQLMR, HIVE and PIG and it has been concluded that E-GENMR shows better performance.