Facebook’s server installation has been a point of interest for a long time. When an engineer on the memcached mailing list casually mentioned that they are running 4TB worth of caches, a lot of people got their panties in a knot.

Since then Facebook has been more forthcoming with details about their exciting architecture. For example they explained how they created a custom Mysql to solve problems with cross-datacenter cache expiry and replication lag which is a great read for anyone looking into multi location hosting.

Today there was another memcached centric post that talks about their fork of the memcached codebase which they also made available on github.

We use more than 800 servers supplying over 28 terabytes of memory to our users. Over the past year as Facebook’s popularity has skyrocketed, we’ve run into a number of scaling issues. This ever increasing demand has required us to make modifications to both our operating system and memcached to achieve the performance that provides the best possible experience for our users. [...]

Fascinating read.