MySQL Versus NewSQL

From our article What is NewSQL, a question possibly will come in mind – MySQL with InnoDB engine is also coming under the classification of NewSQL. Then why we need to think, learn, use the new software such as Google Spanner, Clustrix, NuoDB, TransLattice?

NewSQL is the latest implementation of relational databases. Traditional relational databases including MySQL with MyISAM engine were for a single server which worked well at the older time and still great for many wors. Problem of keeping the database online is obvious issue of MySQL just for a WordPress website. NewSQL DBMS provides fault tolerance and crashes recovery mechanism. But unlike the traditional DBMSs, the newer DBMSs must also minimize downtime. Optimized InnoDB should co-share that advantage.

In real life, it can be hard to scale vertically. Horizontal scaling is cheap in large scale. So, the web and mobile development went towards horizontal scaling technologies – not just the database. You can not use MySQL with InnoDB engine in that easy of way NoSQL. Case of WordPress is different – then web software is already designed. When we are developing software with a proper plan or using a database system for big data, then we need to think around various parameters.

The NoSQL systems have shown that scale-out systems can be done for some specific use case. Hundreds of NewSQL database system exist tailored for specific needs. The NewSQL brought the advantages of ACID and SQL together and provided a simple, realistic scale-out property of NoSQL systems. MySQL with InnoDB is a different journey and progress of the MySQL system. The future will bring more of such systems for various cases. In our older article, we have written: “They are designed to work in a shared-nothing distributed cluster, in which each node has a subset of the data.” NewSQL is the modern RDBMS or it is said to be as the class of SQL database products that handle performance and scalability issues presented by traditional OLTP relational database management systems. Many of the older systems are multi-modal. That means they allow both the traditional SQL RDBMS access methods as well as alternate schema/schemaless options (KV, document store, wide column and so on). Advantages of NewSQL is that they have full transactional support, minimum application complexity, and stronger consistency. They have recognizable SQL and standard tooling and richer analytics. Apache Lucene, Apache Solr and ElasticSearch offered us text and document indexing functions for use case such as real-time search. Graph databases like Neo4J or Titan organize data by relationships instead of by row. With the development of mobile devices, Internet of Things and Big Data we are requiring “out of the box” lighter systems with more use case directed tooling. There are not many ways to classify SAP HANA other than keeping in a broad category. SAP HANA was created as a perfect fit for SAP deployments.

NewSQL is not a single software or a group with too many mandatory common properties. Traditional SQL systems were built one software for all purposes (which includes MySQL). They are good for general purpose applications. Several vendors related to big data including IBM, SAP, Microsoft, Oracle worked behind the NewSQL and still today NewSQL category has debatable matters to the academicians. The NewSQL database systems are not a radical new architecture but rather the continuous development of database technologies.