Redis Data Types Overview
Redis supports 5 basic data types, each with unique characteristics and use cases. This lesson provides an overview of all data types.
Redis Data Types Overview
| Data Type | Description | Internal Implementation | Use Cases |
|---|---|---|---|
| String | Most basic type, stores strings, integers, binary data | SDS (Simple Dynamic String) | Caching, counters, distributed locks |
| Hash | Key-value collection, like Map/Object | Hash table + ziplist | Storing objects, shopping carts |
| List | Ordered, repeatable list of strings | Doubly linked list + ziplist | Message queues, recent lists |
| Set | Unordered, unique string collection | Hash table + intset | Tags, social relationships |
| Sorted Set | Ordered, unique collection with scores per element | Skip list + hash table | Leaderboards, weighted collections |
💡 Why is Redis fast? All data is in memory, and it uses efficient data structures (hash tables, skip lists, etc.).
String
Strings are the most basic Redis data type. They can store:
- Strings (text)
- Integers (numeric)
- Binary data (images, serialized objects)
Features
- Binary safe: can store any data (including images, videos)
- Maximum size: 512MB
- Supports numeric operations: INCR, DECR, INCRBY, etc.
- Supports bit operations: SETBIT, GETBIT, BITCOUNT, etc.
Basic Commands
REDIS
# Set and get
SET key value
GET key
# Numeric operations
INCR key # Increment by 1
INCRBY key 10 # Increment by 10
DECR key # Decrement by 1
# Other operations
APPEND key suffix # Append
STRLEN key # Get length
Use Cases
| Scenario | Example |
|---|---|
| Caching | SET user:1:profile '{"name":"Alice"}' |
| Counter | INCR article:123:views |
| Distributed lock | SET lock:resource "locked" NX EX 10 |
| Session | SET session:token123 "user_data" EX 3600 |
💡 Distributed lock principle: The NX parameter (set only if key doesn't exist) combined with EX (expiration) of the SET command provides atomic lock acquisition.
Hash
A hash is a key-value collection, similar to a Map, Dictionary, or Object in programming languages.
Features
- Ideal for storing objects (user info, product info, etc.)
- Can modify individual fields without reading the entire object
- Memory efficient (uses ziplist for small data)
Basic Commands
REDIS
# Set and get fields
HSET user:1 name "Alice"
HSET user:1 age 25
HGET user:1 name # "Alice"
# Set multiple fields
HMSET user:2 name "Bob" age 30 city "Beijing"
# Get all fields and values
HGETALL user:2
# 1) "name"
# 2) "Bob"
# 3) "age"
# 4) "30"
# 5) "city"
# 6) "Beijing"
# Get multiple fields
HMGET user:2 name age
# 1) "Bob"
# 2) "30"
# Delete a field
HDEL user:2 city
# Check if field exists
HEXISTS user:2 name # 1
# Get all field names
HKEYS user:2
# 1) "name"
# 2) "age"
# Get all values
HVALS user:2
# 1) "Bob"
# 2) "30"
# Get number of fields
HLEN user:2 # 2
Use Cases
| Scenario | Example |
|---|---|
| Storing objects | HSET user:1 name "Alice" age 25 email "alice@example.com" |
| Shopping cart | HSET cart:user1 product:101 2 product:102 1 (product ID and quantity) |
| Counter group | HINCRBY article:123 likes 1 |
Hash vs String for Storing Objects
Method 1: String with JSON
REDIS
SET user:1 '{"name":"Alice","age":25,"email":"alice@example.com"}'
- Pros: Simple
- Cons: Modifying one field requires reading, modifying, and writing back the entire object
Method 2: Hash
REDIS
HSET user:1 name "Alice" age 25 email "alice@example.com"
- Pros: Can modify individual fields independently
- Cons: Doesn't support complex nested structures
💡 Recommendation: Use Hash for simple objects, use String with JSON for complex nested objects.
List
A list is an ordered sequence of strings, allowing duplicate elements.
Features
- Ordered: elements maintain insertion order
- Repeatable: same value can appear multiple times
- Dual-ended operations: can push and pop from both ends
- Max length: 2^32 - 1 (about 4.2 billion)
Basic Commands
REDIS
# Left insert (head)
LPUSH mylist "a" "b" "c" # Returns list length
# Right insert (tail)
RPUSH mylist "d" "e"
# Get list length
LLEN mylist # 5
# Get elements within a range
LRANGE mylist 0 -1 # Get all elements
# 1) "c"
# 2) "b"
# 3) "a"
# 4) "d"
# 5) "e"
# Get element at a specific index
LINDEX mylist 0 # "c" (first element)
# Left pop (head)
LPOP mylist # "c"
# Right pop (tail)
RPOP mylist # "e"
# Set element at a specific index
LSET mylist 0 "new"
# Remove elements by value
LREM mylist 1 "a" # Remove 1 element with value "a"
Use Cases
| Scenario | Description |
|---|---|
| Message queue | LPUSH + RPOP or RPUSH + LPOP |
| Recent list | Latest articles, recent comments |
| Timeline | User's feed, microblog timeline |
Example: Message Queue
BASH
# Producer: push messages to the queue
LPUSH queue:email "send to user1"
LPUSH queue:email "send to user2"
# Consumer: pop messages from the queue
RPOP queue:email
RPOP queue:email
# Blocking pop (blocks when no messages)
BRPOP queue:email 10
💡 LPUSH + RPOP = Queue (FIFO), LPUSH + LPOP = Stack (LIFO)
Set
A set is an unordered collection of unique strings.
Features
- Unordered: elements have no specific order
- Unique: automatic deduplication
- Supports set operations: intersection, union, difference
Basic Commands
REDIS
# Add elements
SADD myset "a" "b" "c"
# Get all elements
SMEMBERS myset
# 1) "a"
# 2) "b"
# 3) "c"
# Check if element exists
SISMEMBER myset "a" # 1 (exists)
# Remove elements
SREM myset "a"
# Get set size
SCARD myset # 2
# Get a random element
SRANDMEMBER myset
# Pop a random element
SPOP myset
Set Operations
REDIS
# Create two sets
SADD set1 "a" "b" "c"
SADD set2 "b" "c" "d"
# Intersection (elements in both sets)
SINTER set1 set2
# 1) "b"
# 2) "c"
# Union (elements in either set)
SUNION set1 set2
# 1) "a"
# 2) "b"
# 3) "c"
# 4) "d"
# Difference (elements in set1 but not in set2)
SDIFF set1 set2
# 1) "a"
# Store result in a new set
SINTERSTORE result set1 set2
Use Cases
| Scenario | Description |
|---|---|
| Tag system | User tags, article tags |
| Social relationships | Friend lists, follow lists |
| Mutual friends | Intersection of two users' friend sets |
| Recommendation system | People you may know = friends of my friends - my friends - me |
Example: Mutual Friends
BASH
# User 1's friends
SADD friends:user1 "Alice" "Bob" "Charlie"
# User 2's friends
SADD friends:user2 "Bob" "Charlie" "David"
# Mutual friends
SINTER friends:user1 friends:user2
Sorted Set
A sorted set is a collection with scores, where elements are sorted by score.
Features
- Ordered: elements sorted by score (ascending)
- Unique: elements are unique, but scores can repeat
- Supports range queries: by score or rank
Basic Commands
REDIS
# Add elements (with score)
ZADD leaderboard 100 "Alice"
ZADD leaderboard 95 "Bob"
ZADD leaderboard 98 "Charlie"
# Get element's score
ZSCORE leaderboard "Alice" # "100"
# Get element's rank (0-based, ascending by score)
ZRANK leaderboard "Alice" # 2 (3rd place)
# Get element's rank (descending by score)
ZREVRANK leaderboard "Alice" # 0 (1st place)
# Get elements by rank range (ascending)
ZRANGE leaderboard 0 -1 WITHSCORES
# 1) "Bob"
# 2) "95"
# 3) "Charlie"
# 4) "98"
# 5) "Alice"
# 6) "100"
# Get elements by rank range (descending)
ZREVRANGE leaderboard 0 -1 WITHSCORES
# 1) "Alice"
# 2) "100"
# 3) "Charlie"
# 4) "98"
# 5) "Bob"
# 6) "95"
# Get elements by score range
ZRANGEBYSCORE leaderboard 95 100 WITHSCORES
# Increment element's score
ZINCRBY leaderboard 5 "Bob" # Bob's score becomes 100
# Remove element
ZREM leaderboard "Bob"
# Get set size
ZCARD leaderboard
# Count elements within a score range
ZCOUNT leaderboard 90 100
Use Cases
| Scenario | Description |
|---|---|
| Leaderboard | Gaming leaderboard, trending list |
| Weighted tags | Tags + weight |
| Delayed queue | Score is execution timestamp |
| Sliding window | Data statistics within a time window |
Example: Gaming Leaderboard
BASH
# Update player scores
ZADD game:leaderboard 1500 "player1"
ZADD game:leaderboard 2300 "player2"
ZADD game:leaderboard 1800 "player3"
# Get top 10
ZREVRANGE game:leaderboard 0 9 WITHSCORES
# Get player rank
ZREVRANK game:leaderboard "player1"
# Get players with scores between 1000-2000
ZRANGEBYSCORE game:leaderboard 1000 2000 WITHSCORES
Data Type Selection Guide
Decision Tree
What do you need to store?
├─ Single value (string, number, binary)
│ └─ Use String
│
├─ Object (multiple fields)
│ ├─ Fields often modified individually
│ │ └─ Use Hash
│ └─ Fields rarely modified individually
│ └─ Use String (JSON)
│
├─ Ordered list
│ ├─ Needs dual-ended operations
│ │ └─ Use List
│ └─ Needs sorting by score
│ └─ Use Sorted Set
│
└─ Unordered collection
├─ Needs deduplication
│ └─ Use Set
└─ Needs deduplication + sorting
└─ Use Sorted Set
Comparison Summary
| Feature | String | Hash | List | Set | Sorted Set |
|---|---|---|---|---|---|
| Ordered | - | - | ✓ | ✗ | ✓ |
| Repeatable | - | - | ✓ | ✗ | ✗ |
| Numeric ops | ✓ | ✓ | ✗ | ✗ | ✓ |
| Set operations | ✗ | ✗ | ✗ | ✓ | ✗ |
| Range queries | ✗ | ✗ | ✓ | ✗ | ✓ |
| Single-field modify | ✗ | ✓ | ✗ | ✗ | ✗ |
TYPE Command
Use the TYPE command to check a key's data type:
REDIS
SET mystring "hello"
TYPE mystring # string
HSET myhash field "value"
TYPE myhash # hash
LPUSH mylist "a"
TYPE mylist # list
SADD myset "a"
TYPE myset # set
ZADD myzset 1 "a"
TYPE myzset # zset
❓ FAQ
Q How much data can a String store?
A Maximum 512MB. However, storing very large data is not recommended as it affects performance.
Q Which is better for storing objects — Hash or String?
A Use Hash for simple objects (individual fields can be modified), use String (JSON) for complex nested objects.
Q How many elements can a List store?
A Up to 2^32-1 (about 4.2 billion), but practically limited by available memory.
Q What is the difference between Set and List?
A Sets are unordered with unique elements; Lists are ordered and allow duplicates. Use Set for deduplication, List for ordered sequences.
Q Can Sorted Set scores be decimals?
A Yes, scores are double-precision floating-point numbers.
📖 Summary
- Redis has 5 basic data types: String, Hash, List, Set, Sorted Set
- String: most basic, stores strings, integers, binary
- Hash: key-value collection, ideal for objects
- List: ordered list, suitable for message queues, recent lists
- Set: unordered collection, supports set operations
- Sorted Set: ordered collection, ideal for leaderboards
- Choose the right data type based on your use case
📝 Exercises
- Create data: Create sample data for each data type
- TYPE command: Use TYPE to verify the data types you created
- Scenario thinking: For an e-commerce site, what data types would you use for product info, shopping cart, and order list?
- Set operations: Create two users' follow lists and calculate their mutual follows
Next Lesson
In the next lesson, we will learn Redis Key Management, diving deeper into key operations and management.