SELECTIVE CONSISTENCY MODEL ON MONGODB BIG DATA STORE

摘要

Data consistency is an integral part of CAP (consistency-availability-partition) guarantees. Data consistency however is rationed to the permissible levels to achieve higher availability, scalability and other performance guarantees. The presented technique is a rationed consistency model applied to selected data objects and is further extensible to selective transactions in NoSQL data store. In this paper, we conceptualize, design and build working prototype of selective data consistency model on MongoDB data store to apply higher levels of consistency only on selected data objects in transactional application and show the improvement in its performance. The paper also contributes by presenting a design of a data criticality analyser for case study on ecommerce application.