What is Memcached?
Memcached is a high-performance, distributed memory caching system designed to speed up dynamic web applications by alleviating database load. It stores data and objects in RAM to reduce the number of times an external data source (such as a database or API) must be read, providing sub-millisecond response times for cached data.
Originally developed by Brad Fitzpatrick for LiveJournal in 2003, Memcached has become one of the most widely used caching solutions in web development. It follows a simple key-value store model and is designed to be simple, fast, and easily scalable across multiple servers. Major companies like Facebook, Twitter, and YouTube rely on Memcached to serve billions of requests daily.
Memcached Performance Calculator
Bandwidth: 80 Kbps
CPU per Node: 7%
Utilization: 7%
Memcached Core Features
Simple Protocol
Text-based protocol that's easy to implement and debug.
• ADD/REPLACE commands
• DELETE operations
• INCREMENT/DECREMENT
• STATS monitoring
High Performance
Optimized for speed with minimal overhead.
• 50K+ requests/second per node
• Event-driven architecture
• Multi-threaded design
• Zero-copy operations
Distributed Design
Client-side distribution for horizontal scaling.
• No single point of failure
• Linear scalability
• Client-side sharding
• Automatic failover
Memory Management
Efficient memory allocation and LRU eviction.
• LRU eviction policy
• Memory pre-allocation
• Configurable item sizes
• Memory usage statistics
Memcached Architecture
Client-Server Model
Simple architecture where clients connect directly to Memcached servers.
[Application] → [Memcached Client Library] → [Memcached Server Pool]
↓
[Server1] [Server2] [Server3]Consistent Hashing
Distributes keys evenly across servers with minimal reshuffling.
• Minimal key migration
• Server addition/removal
• Load balancing
Slab Allocation
Memory management system that reduces fragmentation.
• Different slab classes
• Efficient allocation
• Reduced fragmentation
Memcached Use Cases
Database Query Caching
Cache expensive database query results to reduce load.
- • Frequently accessed data
- • Complex query results
- • Aggregated statistics
- • User profile data
Session Storage
Store user session data for web applications.
- • User authentication tokens
- • Shopping cart contents
- • User preferences
- • Temporary form data
API Response Caching
Cache external API responses to improve performance.
- • Third-party API responses
- • Rate-limited API calls
- • Expensive computations
- • Static content
Content Delivery
Cache rendered content and static assets.
- • Rendered HTML pages
- • Image metadata
- • CSS/JS fragments
- • Template output
Memcached Commands
Basic Operations
# Store data
set key 0 3600 5
hello
STORED
# Retrieve data
get key
VALUE key 0 5
hello
END
# Delete data
delete key
DELETEDStorage Commands
• add - Store if not exists
• replace - Store if exists
• append - Append to existing
• prepend - Prepend to existing
• cas - Check and set
Retrieval Commands
• gets - Get with CAS value
• mget - Multiple key retrieval
• gat - Get and touch (refresh TTL)
• touch - Update expiration time
Numeric Operations
• decr - Decrement numeric value
• Atomic operations
• 64-bit unsigned integers
• Overflow protection
Administrative
• flush_all - Clear all data
• version - Server version
• quit - Close connection
• slabs - Slab information
Real-World Memcached Implementations
Massive Memcached deployment serving billions of requests daily.
- • 28TB of cached data
- • Millions of requests per second
- • Custom optimizations
- • Multi-tier architecture
Wikipedia
Caches rendered articles and database queries for fast page loads.
- • Article content caching
- • User session storage
- • Database query results
- • Multi-language support
YouTube
Uses Memcached for video metadata and user data caching.
- • Video metadata
- • User preferences
- • View count data
- • Recommendation cache
Caches post data, comments, and user sessions for fast response times.
- • Post and comment data
- • User authentication
- • Vote counting
- • Subreddit metadata
Memcached Best Practices
✅ Do
- • Use consistent key naming conventions
- • Set appropriate expiration times
- • Monitor cache hit ratios
- • Use connection pooling
- • Implement proper error handling
- • Consider using binary protocol
- • Monitor memory usage and stats
❌ Don't
- • Store critical data without backup
- • Use Memcached as a persistent store
- • Ignore cache warming strategies
- • Store large objects (>1MB)
- • Forget to handle cache misses
- • Use predictable key patterns
- • Neglect security considerations