SoulEvilX

Both tools are powerful, fast, in-memory data stores that are useful as a cache. Both can help speed up your application by caching database results, HTML fragments, or anything else that might be expensive to generate.

Points to Consider

When used for the same thing, here is how they compare using the original question’s “Points to Consider”:

• Read/write speed: Both are extremely fast. Benchmarks vary by workload, versions, and many other factors but generally show redis to be as fast or almost as fast as memcached. I recommend redis, but not because memcached is slow. It’s not.
• Memory usage: Redis is better.
  • memcached: You specify the cache size and as you insert items the daemon quickly grows to a little more than this size. There is never really a way to reclaim any of that space, short of restarting memcached. All your keys could be expired, you could flush the database, and it would still use the full chunk of RAM you configured it with.
  • redis: Setting a max size is up to you. Redis will never use more than it has to and will give you back memory it is no longer using.
  • I stored 100,000 ~2KB strings (~200MB) of random sentences into both. Memcached RAM usage grew to ~225MB. Redis RAM usage grew to ~228MB. After flushing both, redis dropped to ~29MB and memcached stayed at ~225MB. They are similarly efficient in how they store data, but only one is capable of reclaiming it.
• Disk I/O dumping: A clear win for redis since it does this by default and has very configurable persistence. Memcached has no mechanisms for dumping to disk without 3rd party tools.
• Scaling: Both give you tons of headroom before you need more than a single instance as a cache. Redis includes tools to help you go beyond that while memcached does not.

Continue reading “Memcached vs Redis: Direct Comparison”