Memcached vs Redis: Which In‑Memory Cache Wins for Your Applications?

1. Memcached and Redis

When handling large volumes of data, caching is commonly used to improve performance. Memcached is a simple, distributed memory cache system ideal for caching or session storage, while Redis is an in‑memory data‑structure store offering a rich set of features and can serve as a cache, database, message broker, and queue.

2. Installation

2.1 Install Memcached

Download the latest Memcached package and compile it:

$ wget http://memcached.org/latest
$ tar -zxvf memcached-1.6.3.tar.gz
$ cd memcached-1.6.3
$ ./configure && make && make test && sudo make install

2.2 Install Redis

Similarly, install the latest Redis server:

$ wget http://download.redis.io/releases/redis-5.0.8.tar.gz
$ tar xzf redis-5.0.8.tar.gz
$ cd redis-5.0.8
$ make

3. Similarities

3.1 Sub‑millisecond latency

Both Memcached and Redis store data in memory, providing sub‑millisecond response times.

3.2 Data partitioning

Both support distributing data across multiple nodes.

3.3 Language support

Both work with major programming languages such as Java, Python, JavaScript, C, and Ruby, and have client libraries (e.g., Xmemcached, Memcached‑java‑client for Memcached; Jedis, Lettuce, Redisson for Redis).

3.4 Cache clearing

Memcached uses flush_all to clear the cache, while Redis provides FLUSHDB and FLUSHALL .

3.5 Scalability

Both solutions scale well to handle exponential growth in demand.

4. Differences

4.1 Command‑line interface

Memcached can be accessed via telnet:

$ telnet 192.168.8.123 6605
stats
...

Redis offers its own CLI, redis-cli, for executing commands.

4.2 Disk I/O persistence

Memcached relies on third‑party tools for disk dumps, whereas Redis provides built‑in mechanisms (RDB snapshots and AOF logs) for persistence.

4.3 Data structures

Memcached stores simple string key‑value pairs (max 1 MB per value). Redis supports richer structures such as lists, sets, hashes, and can store values up to 512 MB.

4.4 Replication

Memcached can use third‑party tools like repcached for replication. Redis has native master‑replica replication using REPLICAOF and PSYNC.

4.5 Transactions

Memcached does not support transactions (operations are atomic). Redis provides transactional commands: MULTI, EXEC, and conditional WATCH.

4.6 Pub/Sub messaging

Redis includes built‑in publish/subscribe commands ( PUBLISH, SUBSCRIBE, UNSUBSCRIBE), useful for real‑time communication; Memcached lacks this feature.

4.7 Geospatial support

Redis offers geospatial commands such as GEODIST and GEORADIUS for location‑based queries, which Memcached does not provide.

4.8 Architecture

Memcached is multithreaded and can leverage multiple CPU cores, giving it an advantage for large data sets. Redis is single‑threaded but can be scaled horizontally via clustering.

4.9 Lua scripting

Redis allows execution of Lua scripts using EVAL and SCRIPT LOAD, enabling complex atomic operations.

4.10 Memory efficiency

For simple string storage, Memcached uses memory more efficiently, but Redis achieves higher efficiency when using hash structures.

5. Conclusion

Memcached is a reliable choice for straightforward caching needs. However, Redis offers a richer feature set—including advanced data structures, persistence, replication, transactions, pub/sub, geospatial queries, and scripting—making it better suited for complex scenarios.