In 2011, Matthew Aslett coined the term "NewSQL" to represent an emerging category of high-performance relational database management systems (DBMSs) for online transaction processing (OLTP) workloads. NewSQL systems were in response to the NoSQL movement that argued that DBMSs should eschew key elements of relational DBMSs (e.g., transactions, joins, SQL) in favor of performance and availability.
Ten years later, the database landscape has changed. Most of the original NewSQL DBMSs still exist today, but newer systems have taken over their mindshare, or they pivoted to other application domains. The NoSQL design philosophy has also wanned as prominent systems in this group have adopted NewSQL principles.
In this talk, we will first discuss the origins of the NewSQL movement and its influence on the DBMS marketplace. We will cover the systems that were part of the original NewSQL classification and then how both the category and those systems evolved. This discussion seeks to address whether NewSQL is a meaningful way to understand a system's capabilities or whether it is just marketing fluff. The second part of this talk will cover whether NewSQL is still a relevant term today. We will also consider what OLTP DBMSs might look like in the upcoming decade.