Scalable in-memory data management model for enterprise applications

Abstract

With the rapid advances in technology and data volume, having efficient and scalable data management system is essential for most of the enterprise applications. So In-Memory data management systems are becoming the highly used data management solution in most of the time critical enterprise solutions. Although In Memory Data Management Systems are widely used, still they are having problems such as scalability issues, concurrency problems etc. This project is an effort that aims to propose a scalable enterprise solution for in memory data management, identifying the bottlenecks in the current In-Memory Data management systems. Although there are various benchmarks are available for Disk Resident Databases, lack of a fair metric for comparing the performance of different in-memory database systems has become a problem when selecting the appropriate data management system for enterprise applications. Currently there are various in-memory databases are available and when using them with the enterprise applications, developers have to put lot of effort as there is no standard API/Interfaces available for them. This research project addresses these two problems by providing an unbiased performance benchmark for various in-memory databases and developing a data connector framework to access different data sources such as in-memory databases, disk resident databases, flat file data bases and in-memory data caches. This report provides details about the problem background, existing system implementations and current research areas in this domain and how I’m going to achieve the objective.