Course contentsShow
← System Design
Lesson 331 of 1,9199. SQL vs NoSQL TradeoffsFree lesson

What NewSQL Is

NewSQL systems aim to provide NoSQL-like scalability while maintaining ACID guarantees and SQL interfaces.

What NewSQL Is

You'll learn: What NewSQL databases are and how they bridge the gap between traditional SQL and NoSQL systems.

The Best of Both Worlds

Remember how we've been exploring the ACID vs BASE tradeoff? Traditional SQL databases give you strong consistency and transactions but struggle with horizontal scaling. NoSQL databases scale horizontally but sacrifice ACID guarantees. NewSQL systems aim to solve this dilemma by offering both NoSQL-style scalability and traditional SQL capabilities with ACID guarantees.

The Core Idea

Think of NewSQL as a modern reimagining of relational databases. These systems recognize that the core SQL model and ACID properties are valuable, but the underlying architecture of traditional databases (designed decades ago) wasn't built for today's distributed, cloud-native world.

NewSQL databases use innovative distributed architectures—similar to what NoSQL pioneered—while maintaining:

  • Full SQL interface for queries and schema definition
  • ACID transactions across distributed nodes
  • Horizontal scalability by adding more machines
  • Strong consistency guarantees

They essentially say: "You shouldn't have to choose between consistency and scale. Modern distributed systems techniques let us have both."

How They Differ

Unlike traditional SQL databases that scale vertically (bigger machines), NewSQL scales horizontally (more machines). Unlike NoSQL, you keep your familiar SQL queries, joins, and guaranteed transactions—no eventual consistency compromises required.

Examples include Google Spanner, CockroachDB, and VoltDB, though you'll learn the specifics of individual systems later.

Key Takeaway: NewSQL databases combine NoSQL's horizontal scalability with SQL's ACID guarantees and relational model, offering a middle ground for applications that need both consistency and distributed scale.