Memcached
Memcached is an open-source, high-performance, distributed in-memory key-value store used primarily to cache the results of database queries, API calls, and page rendering to reduce load on backend systems.
Definition
Memcached is an open-source, high-performance, distributed in-memory key-value store used primarily to cache the results of database queries, API calls, and page rendering to reduce load on backend systems.
Overview
Memcached was originally created to speed up LiveJournal by caching the output of expensive database queries in RAM instead of recomputing them on every request. It exposes a simple key-value protocol: applications set a value under a key with an optional expiration time, and later fetch it back with a lookup, avoiding a round trip to a slower system of record such as MySQL or PostgreSQL. Architecturally, Memcached is deliberately minimal. It stores data purely in memory with no persistence to disk, uses a simple slab allocator to manage memory efficiently, and scales horizontally by having client libraries hash keys across a pool of Memcached servers, called sharding. If a node fails, its cached data is simply lost and can be regenerated from the source of truth, which keeps the system simple at the cost of durability. Memcached is frequently compared to Redis, which offers richer data structures, persistence options, and built-in replication, whereas Memcached focuses narrowly on being a fast, multithreaded cache with a small, predictable feature set. Many web applications and content platforms still choose Memcached specifically for its simplicity and multithreaded performance under high concurrent read load, often placed in front of a data warehouse or primary database as part of a broader caching layer.
Key Features
- Simple, fast key-value get/set/delete protocol
- Pure in-memory storage with no disk persistence
- Multithreaded architecture for high request throughput
- Client-side sharding across a pool of cache servers
- Automatic eviction of least-recently-used items when memory is full
- Language-agnostic client libraries for most major programming languages
- Minimal operational overhead due to its stateless, cache-only design