Redis Basics

Redis is an in-memory data store — it keeps everything in RAM, making it extremely fast (sub-millisecond responses). This is why companies use it for caching, session storage, rate limiting, real-time leaderboards, and pub/sub messaging. Unlike traditional databases that read from disk, Redis trades durability for speed — perfect when you need instant answers. In this lesson, you connect to Redis and verify it is running.

Strings are Redis's simplest data type: one key, one value. But with SETEX (SET + EXpiry), they become powerful for real-world use cases. Every website that remembers your login uses something like this — session tokens stored with a TTL that auto-expire. Rate limiters count requests per IP with expiring keys. Caches store expensive database query results for a few minutes. The pattern is always the same: SET a key, let Redis auto-delete it after N seconds.

Redis Lists are ordered collections that support push/pop from both ends — making them natural queues. In production, they power job queues (workers RPOP tasks that producers LPUSH), activity feeds (LPUSH new events, LRANGE to display recent ones), and message buffers. Unlike a database table, a Redis list handles thousands of push/pop operations per second with zero overhead. The key insight: if you need ordered, fast, temporary data — use a list.

Redis Sets store unique, unordered elements — no duplicates allowed. This makes them perfect for problems where uniqueness matters: tracking unique visitors (SADD each visitor, SCARD to count), tagging systems (each tag is a set of items), and social features like mutual friends (SINTER to find common connections). Set operations (union, intersection, difference) happen server-side in microseconds, which would take multiple queries in a traditional database.

Redis Hashes store multiple field-value pairs under one key — like a mini-object or row. In production, they are used for user profiles (one hash per user with fields like name, email, role), shopping carts (product IDs as fields, quantities as values), and configuration stores. The advantage over Strings: you can read or update a single field (HGET, HSET) without fetching the whole object. This makes Hashes memory-efficient and fast for structured data that changes field by field.