What do you like best about the product?

I like that Memcached allows for storage in RAM. It is Key-Value, but storing a JSON string is easy and quick. Especially so when using for it's intended purpose as a cache without worrying about persistence. It's written in C, very close to the metal. Additionally, development on it is very active - with the most recent stable release a few months ago. I especially love that it can distribute its hash table across multiple machines, thus improving speed and reducing overall workload.

What do you dislike about the product?

While the prefered cleanup strategy is LRU (Least Recently Used) I would love to have the ability to change to LFU (Least Frequently Used) for instance. I am not aware of this being configurable. Care should also be taken to make sure that it is secure (typically user error). One should configure the application correctly and to only accept connections from whitelisted applications.

What problems is the product solving and how is that benefiting you?

I use it primarily purely as a cache, no persistence. It sped up our load times and user lookup for subsequent visits and page loads dramatically, in the order of 300% to 400% faster load times.

Recommendations to others considering the product:

Make sure you tie up security in the config before going live.