Evaluation of ACE Properties of Traditional SQL and NoSQL Big Data Systems

Traditional SQL and NoSQL big data systems are the backbone for managing data in cloud, fog and edge computing. This paper develops a new system and adopts the TPC-DS industry standard benchmark in order to evaluate three key properties, availability, consistency and efficiency (ACE) of SQL and NoSQL systems. The contributions of this work are manifold. It evaluates and analyses the tradeoff between the ACE properties. It provides insight into the NoSQL systems and how they can be improved to be sustainable for a more wide range of applications. The evaluation shows that SQL provides stronger consistency, but at the expense of low efficiency and availability. NoSQL provides better efficiency and availability but lacks support for stronger consistency. In order for NoSQL systems to be more sustainable they need to implement transactional schemes that enforce stronger consistency as well as better efficiency and availability.