A basic overview of the standard "best practices" server architecture for Drupal (the "Mercury stack") will be covered, as well as extending it to clusters of servers. We will discuss evolutions along the lines of memcache vs. redis and Apache vs. Nginx. Finally, some of the current "open problems" for which there is no commonly accepted solution are described - clustered file systems, high write-traffic leading to "NoSQL" type databases, etc.

The material tends more towards the underlying server, OS, database, and etc, rather than Drupal configuration itself.

This presentation will not hand you any magic bullets, but you will have a basis for making decisions and enough knowledge to ask good questions about your own scaling strategy.