When to Use Each Redis Data Structure: Real‑World Scenarios Explained

Redis Data Structure Use Cases

Having read Redis books and explored its basic commands, I started examining the source code to master Redis data structures. Redis provides five fundamental structures, each suited to specific scenarios.

String – simple key‑value pairs

Hash – dictionary‑like objects

List – ordered collections

Set – unique unordered collections

Sorted Set – ordered collections with scores

Below is a brief description of typical applications for each.

1. String – simple values

Strings store plain text or numbers and can replace Memcached with higher efficiency, supporting persistence, replication, and operations like LEN, APPEND, substring get/set, bit‑level manipulation, bulk set, atomic counters, and GETSET for atomic replace.

2. Hash – field‑value maps

Instead of serializing a whole JSON object in Memcached, Redis Hashes let you update a single field directly, ideal for user attributes (nickname, age, gender, points) without costly read‑modify‑write cycles.

Store, read, and modify individual user properties

3. List – linked‑list collections

Implemented as a doubly linked list, Lists enable timelines (e.g., Twitter feed) and message queues: push tasks onto the list and workers pop them for processing. Redis also provides range queries and deletions within a List.

Timeline feeds

Message queues

4. Set – unique element collections

Sets hold distinct values and support operations like intersection, union, and difference, useful for social features such as followers/friends, counting unique IPs, or recommending users based on common tags.

Mutual friends, second‑degree connections

Unique visitor counting

Tag‑based friend recommendations

5. Sorted Set – ordered sets with scores

Sorted Sets add a numeric score to each member, enabling natural ordering. Common uses include leaderboards (score = user points) and weighted queues where higher‑score items are processed first.

Ranked leaderboards

Weighted task queues

Other Redis Features and Their Scenarios

1. Pub/Sub – real‑time messaging

Publish/Subscribe lets clients listen to messages on a key; when a message is published, all subscribers receive it instantly, making it suitable for chat systems, group messaging, and other real‑time notifications.

2. Transactions – command batching

Redis transactions group commands for atomic execution (though not full ACID). The WATCH command can monitor keys and abort the transaction if a watched key changes, providing optimistic concurrency control.