Modeling and enforcing integrity constraints on graph databases

Abstract

The enormous volume and high variety of information that is constantly produced by computing systems requires storage technologies able to provide high processing velocity and data quality. The suitability for modeling complex data and for delivering performance are characteristics that are making graph databases become very popular. However, existing limitations still prevent database management systems that adopt the graph model to fully ensure data consistency, given that the means for ensuring data consistency are usually nonexistent or at most very simple. This work intends to overcome this limitation by extending the support for defining and enforcing integrity constraints on graph databases, in order to prevent the graph to reach an inconsistent state and compromise the correctness of applications. The proposed integrity constraints are implemented on OrientDB. Experimental results show that the prototype implementation can improve the performance in comparison to verification of constraints on a client application.