Redis Hashes (Part 1)
A hash is a key-value collection, ideal for storing objects. This lesson covers basic hash operations.
What is a Hash?
A hash is similar to:
- Python's dict
- JavaScript's Object
- Java's HashMap
A single Redis hash key can store multiple field-value pairs:
user:1
├─ name → "Alice"
├─ age → "25"
├─ city → "Beijing"
└─ email → "alice@example.com"
💡 Advantage: Hashes allow you to modify individual fields without reading the entire object. They are also more memory-efficient than storing JSON strings.
HSET and HGET: Set and Get Fields
HSET: Set a Single Field
REDIS
# Set a single field
HSET user:1 name "Alice"
(integer) 1 # Returns the number of newly added fields
# Set multiple fields
HSET user:1 age 25 city "Beijing"
(integer) 2 # Added 2 new fields
# Update an existing field
HSET user:1 name "Bob"
(integer) 0 # Field already exists, returns 0
HGET: Get a Single Field
REDIS
HGET user:1 name
"Alice"
HGET user:1 age
"25"
# Field does not exist
HGET user:1 notexist
(nil)
HSETNX: Set Only if Field Does Not Exist
REDIS
# Set only if the field does not exist
HSETNX user:1 name "Charlie"
(integer) 0 # Field already exists, failed
HSETNX user:1 phone "13800138000"
(integer) 1 # Field does not exist, set successfully
HMSET and HMGET: Batch Operations
HMSET: Batch Set Fields
REDIS
HMSET user:2 name "Bob" age 30 city "Shanghai" email "bob@example.com"
OK
⚠️ Note: After Redis 4.0.0, HMSET is deprecated. Use HSET for setting multiple fields instead.
HMGET: Batch Get Fields
REDIS
HMGET user:2 name age city
1) "Bob"
2) "30"
3) "Shanghai"
# Including non-existent fields
HMGET user:2 name email phone
1) "Bob"
2) "bob@example.com"
3) (nil) # phone does not exist
HGETALL: Get All Fields and Values
REDIS
HGETALL user:2
1) "name"
2) "Bob"
3) "age"
4) "30"
5) "city"
6) "Shanghai"
7) "email"
8) "bob@example.com"
⚠️ Note: HGETALL returns an alternating array of field names and values. If the hash is large, this uses significant memory and network bandwidth — use with caution!
HKEYS and HVALS: Get All Field Names or Values
HKEYS: Get All Field Names
REDIS
HKEYS user:2
1) "name"
2) "age"
3) "city"
4) "email"
HVALS: Get All Values
REDIS
HVALS user:2
1) "Bob"
2) "30"
3) "Shanghai"
4) "bob@example.com"
HEXISTS: Check if Field Exists
REDIS
HEXISTS user:2 name
(integer) 1 # Exists
HEXISTS user:2 phone
(integer) 0 # Does not exist
HDEL: Delete Fields
REDIS
# Delete a single field
HDEL user:2 email
(integer) 1 # Deleted 1 field
# Delete multiple fields
HDEL user:2 age city
(integer) 2 # Deleted 2 fields
# Delete a non-existent field
HDEL user:2 notexist
(integer) 0
HLEN: Get Number of Fields
REDIS
HSET user:3 name "Charlie" age 28
HLEN user:3
(integer) 2 # Has 2 fields
# Key does not exist
HLEN notexist
(integer) 0
HSTRLEN: Get Field Value Length
REDIS
HSET user:1 name "Alice" bio "Software Engineer"
HSTRLEN user:1 name
(integer) 5 # "Alice" has length 5
HSTRLEN user:1 bio
(integer) 17 # "Software Engineer" has length 17
Hash Use Cases
Use Case 1: Storing User Info
REDIS
# Store basic user info
HSET user:1001 name "Alice" age 25 email "alice@example.com" role "admin"
# Get user name
HGET user:1001 name
# Update user age
HSET user:1001 age 26
# Get all user info
HGETALL user:1001
Use Case 2: Shopping Cart
REDIS
# User 1's shopping cart
HSET cart:user:1 product:101 2 # Product 101, quantity 2
HSET cart:user:1 product:102 1 # Product 102, quantity 1
HSET cart:user:1 product:103 5 # Product 103, quantity 5
# View all items in cart
HGETALL cart:user:1
# Update item quantity
HSET cart:user:1 product:101 3
# Remove an item
HDEL cart:user:1 product:102
# Check how many items in cart
HLEN cart:user:1
Use Case 3: Article Metadata
REDIS
# Store article info
HSET article:123 title "Redis Tutorial" author "Alice" views 1000 likes 50
# Increment view count (requires HINCRBY)
HINCRBY article:123 views 1
# Increment like count
HINCRBY article:123 likes 1
# Get article title and author
HMGET article:123 title author
Use Case 4: Product Cache
REDIS
# Cache product details
HSET product:2001 name "iPhone 15" price 5999 stock 100 category "phone"
# Get product price
HGET product:2001 price
# Update stock
HSET product:2001 stock 95
# Check if product exists
HEXISTS product:2001 name
Hash vs String for Storing Objects
Method 1: String with JSON
REDIS
SET user:1 '{"name":"Alice","age":25,"email":"alice@example.com"}'
# Get the whole object
GET user:1
# Modify age: need to read, parse, modify, serialize, write back
# Multiple steps, inefficient
Method 2: Hash
REDIS
HSET user:1 name "Alice" age 25 email "alice@example.com"
# Get a single field
HGET user:1 name
# Modify age: one-step operation
HSET user:1 age 26
# Get multiple fields
HMGET user:1 name age
Comparison Summary
| Aspect | String (JSON) | Hash |
|---|---|---|
| Modify single field | Need to read the whole object | Direct modification |
| Memory usage | Larger (JSON format overhead) | Smaller (compact storage) |
| Query single field | Need to parse JSON | Direct access |
| Complex nested objects | ✅ Supported | ❌ Not supported |
| Partial update | ❌ Difficult | ✅ Simple |
💡 Selection guide:
- Simple objects (flat structure): use Hash
- Complex objects (nested structure): use String with JSON
- Frequently modifying individual fields: use Hash
Hash Encoding Optimization
Redis automatically selects encoding based on hash size:
| Encoding | Condition | Description |
|---|---|---|
| ziplist | Fields <= 512, value length <= 64 bytes | Compressed list, saves memory |
| hashtable | Fields > 512 or value length > 64 bytes | Hash table, better performance |
Checking Encoding Type
REDIS
HSET small:hash field1 "value1" field2 "value2"
OBJECT ENCODING small:hash
"ziplist"
# After adding many fields
HSET large:hash field1 "value1" field2 "value2" ... field600 "value600"
OBJECT ENCODING large:hash
"hashtable"
ℹ️ Note: Encoding is automatically selected by Redis. You can adjust thresholds through config parameters.
❓ FAQ
Q How many fields can a hash store?
A Theoretically unlimited (2^32 - 1), but practically limited by memory. It's recommended to keep individual hashes under a few thousand fields.
Q Does HGETALL affect performance?
A Yes. If the hash is large, HGETALL returns a lot of data, consuming memory and bandwidth. Use HSCAN for iterative retrieval.
Q Can hash field values be numeric?
A Yes. All values in Redis are strings, but numeric strings support arithmetic operations (HINCRBY).
Q How do I delete an entire hash key?
A Use DEL to delete the key, not HDEL to delete fields:
DEL user:1.Q Does hash support nesting?
A No. Hash values can only be strings. For nesting, use String with JSON.
📖 Summary
- Hash is a key-value collection, ideal for storing objects
- HSET/HGET to set and get individual fields
- HMSET/HMGET for batch field operations
- HGETALL to get all fields and values (use with caution)
- HKEYS/HVALS to get all field names or values
- HEXISTS checks field existence, HDEL deletes fields, HLEN gets field count
- Hashes are great for storing user info, shopping carts, article metadata
- Use Hash for simple objects, String JSON for complex nested objects
📝 Exercises
- User info: Use a hash to store user info (name, age, email, city), then retrieve all info
- Shopping cart: Implement a shopping cart with a hash — add 3 products, modify 1 quantity, delete 1 product
- Field operations: Use HEXISTS to check field existence, HLEN to count fields
- Comparison practice: Store the same object with both String JSON and Hash, compare the steps to modify a single field
Next Lesson
In the next lesson, we will learn Redis Hashes (Part 2), covering advanced hash operations and numeric calculations.