Non-Relational vs. Relational Databases

These flashcards review the critical distinctions between relational and non-relational (including NoSQL) databases, their key properties, use cases, scaling strategies, and the four principal types of non-relational data stores.

What is the fundamental structural difference between relational and non-relational databases?

Relational databases organize data in tables with fixed schemas and relationships, whereas non-relational databases use flexible data models such as key-value, document, columnar, or graph without a rigid tabular structure.

Are the terms “non-relational” and “NoSQL” synonymous?

No. NoSQL databases are a subset of non-relational databases designed for scalability, flexibility, and performance, but some non-relational systems fall outside the NoSQL category.

Which four properties are summarized by the ACID acronym?

Atomicity, Consistency, Isolation, and Durability.

How do relational databases typically scale compared to non-relational databases?

Relational databases usually scale vertically by upgrading hardware, while non-relational databases are designed to scale horizontally across many servers or clusters.

Why can schema changes cause downtime in relational databases but not usually in non-relational databases?

Relational schemas are rigid and must be predefined, so changes often require migrations; non-relational databases allow on-the-fly schema changes due to flexible data models.

Which query language is standard for relational databases?

SQL (Structured Query Language).

Why might non-relational query languages be less expressive than SQL?

They are optimized for the specific data model (e.g., key-value, document) rather than for complex joins and aggregations typical of SQL.

Name two business uses where a relational database is typically preferred.

E-commerce order processing and financial systems (or inventory management).

Name two business uses where a non-relational database is typically preferred.

Real-time analytics and social media/message platforms (e.g., Facebook Messenger).

What was Amazon’s reason for choosing a NoSQL database for product pages?

They needed to serve about 10,000 pages per second with easy horizontal scalability for over a billion products.

When working with unstructured or semi-structured data that lacks clear tables and relationships, which database type is usually recommended?

A non-relational (NoSQL) database.

List the four main types of non-relational databases.

Document-oriented, Key-Value Store, Wide-Column Store, and Graph Store.

How does a document-oriented database store information?

It pairs a key with a complex data structure (similar to a document) that can hold nested or varied fields.

Give a real-world example application well-suited for a document-oriented database.

Messaging applications like Facebook Messenger that store varied, unstructured message content.

What is the defining characteristic of a key-value store?

Each key maps to exactly one value, offering a very simple and fast lookup model.

Which everyday operating-system component uses a key-value store model?

The Windows Registry for storing system settings.

How do wide-column stores differ from traditional relational tables?

Columns can vary in name and format from one row to the next within the same table, giving per-row flexibility.

Describe the data structure used by graph stores.

They link data using nodes, edges, and properties to represent complex relationships.

What advantage do non-relational databases often gain by relaxing strict ACID compliance?

Higher performance and greater scalability for large, distributed workloads.

Define ‘Schema’ in the context of databases.

The overall structure of a database, including its tables (or collections) and the relationships between them.

Define ‘Non-Relational Database.’

Any database that does not rely on the traditional tabular model of relational database management systems.

Why are indexes handled differently in relational versus many non-relational databases?

Relational databases often build and manage indexes internally for efficient queries, whereas non-relational systems may rely more on application-level logic or simpler index mechanisms tailored to their data model.