16 Essential Redis Patterns Every Backend Engineer Should Master

This article presents 16 practical Redis patterns for building scalable backend services, including caching, distributed sessions, distributed locks, global ID generation, counters, rate limiting, bitmap statistics, shopping cart, user timeline, message queue, lottery, likes/check‑in/clock‑in, product tags, product filtering, follow/fan relationships, and leaderboards.

Cache

Distributed data sharing (session store)

Distributed lock

Global ID

Counter

Rate limiting

Bitmap statistics

Shopping cart

User timeline

Message queue

Lottery

Like / Check‑in / Clock‑in

Product tags

Product filtering

User follow / recommendation model

Leaderboard

1. Cache

Store hot data such as report results, popular items, or full‑page HTML in Redis strings. Reads are served from memory, dramatically reducing latency.

2. Distributed Data Sharing (Session Store)

Redis runs as a standalone distributed service, so multiple applications can share data. A common use‑case is a distributed session store. Example Maven dependency for Spring Session with Redis (escaped):

<dependency>
  <groupId>org.springframework.session</groupId>
  <artifactId>spring-session-data-redis</artifactId>
</dependency>

3. Distributed Lock

Implement a lock with the SETNX command. If the key does not exist the command succeeds; set an expiration to avoid deadlocks.

public static boolean getLock(String key) {
    Long flag = jedis.setnx(key, "1");
    if (flag == 1) {
        jedis.expire(key, 10); // lock expires after 10 seconds
    }
    return flag == 1;
}

public static void releaseLock(String key) {
    jedis.del(key);
}

4. Global ID

Use the atomic integer command INCRBY to generate sequential IDs across the cluster, e.g. INCRBY userid 1000 to allocate a block of IDs for sharding.

5. Counter

Simple integer counters with INCR can track page views, likes, etc. The value can be periodically persisted to a relational database.

6. Rate Limiting

Key each client (e.g., IP) with an integer counter. Increment on each request and reject when the counter exceeds a predefined threshold within a time window.

7. Bitmap Statistics

Redis bitmap commands ( SETBIT, BITCOUNT) store one‑bit flags efficiently. Example: track online users.

SET k1 a
SETBIT k1 6 1
SETBIT k1 7 0
GET k1

To find users online every day for a week, use BITOP AND on the daily bitmap keys.

BITOP AND 7_days_both_online_users day_1_online_users day_2_online_users ... day_7_online_users

8. Shopping Cart

Store a cart in a Redis HASH where key = userId, field = productId, value = quantity. Common operations: HINCRBY – add or increase quantity HDEL – remove a product HGETALL – list all items HLEN – count distinct products

9. User Timeline

Use a Redis LIST (doubly linked list) to store ordered timeline entries. New entries are pushed with LPUSH or RPUSH, preserving order.

10. Message Queue

Redis lists provide blocking pop operations BLPOP and BRPOP with a timeout, enabling a reliable FIFO queue. BLPOP key timeout – removes and returns the first element, blocking if the list is empty. BRPOP key timeout – removes and returns the last element, blocking if the list is empty.

11. Lottery

Use a Redis SET and the SPOP command to draw a random element.

SPOP myset

12. Like / Check‑in / Clock‑in

Model likes with a set keyed by like:postId. Example commands: SADD like:t1001 u3001 – add a like SREM like:t1001 u3001 – remove a like SISMEMBER like:t1001 u3001 – test if a user liked SMEMBERS like:t1001 – list all users who liked SCARD like:t1001 – count likes

13. Product Tags

Maintain tags for a product with a set, e.g. SADD tags:i5001 "clear display". Each tag is a separate member.

14. Product Filtering

Combine attribute sets with set operations to filter products. Example: intersect brand, OS, screen size range, and screen type.

# add attributes
SADD brand:apple iPhone11
SADD brand:ios iPhone11
SADD screensize:6.0-6.24 iPhone11
SADD screentype:lcd iPhone11

# filter
SINTER brand:apple brand:ios screensize:6.0-6.24 screentype:lcd

15. User Follow / Recommendation Model

Represent follow relationships with two sets per user: follow:userId (users this user follows) and fans:userId (users who follow this user). Set operations enable mutual follows, suggestions, etc.

SADD 1:follow 2

SADD 2:fans 1

SINTER 1:follow 2:fans

– users I follow who also follow me (mutual). SDIFF 2:follow 1:follow – users followed by user 2 but not by user 1 (potential acquaintances).

16. Leaderboard

Use a sorted set ( ZSET) to maintain scores. Increment a member’s score with ZINCRBY and retrieve the top N with ZREVRANGE … WITHSCORES.

ZINCRBY hotNews:20190926 1 n6001   # increase click count
ZREVRANGE hotNews:20190926 0 14 WITHSCORES   # top 15 items