Introduction to Databases

Introduction

Every backend stores state somewhere. SQL databases (PostgreSQL, MySQL) remain the default for transactional systems because of ACID guarantees, mature tooling and decades of operational experience.

Understanding the topic

SQL vs NoSQL — pick by access pattern, not hype:

• Relational (SQL) — strong schema, joins, transactions. Default choice for OLTP.
• Document (Mongo) — flexible schema, denormalised reads. Good for catalogs, CMS.
• Key-Value (Redis) — sub-ms reads. Cache, sessions, rate-limits.
• Search (Elasticsearch) — full-text, analytics over billions of docs.
• Time-series (Timescale, Influx) — metrics, IoT.

Real-world use

A typical e-commerce backend uses Postgres for orders/users (ACID), Redis for sessions and cart cache, Elasticsearch for product search, Kafka for event log, S3 for images.