The social network MySpace has about 24 billion page views per month, Facebook now includes more than 300 million users, daily more than 35 million people will update their personal information and in the same period more than 40 million new photos are added its database.
This is not only the case of Facebook and MySpace with the larger social networks that exist today operate normally million accounts, up to date with very high frequency. Performance that operators must continue to meet the users are so impressive, if related to the volume of traffic and the needs of users.

Facebook, MySpace and Linkedln using open source software for their functions, based on Linux for the operation of their departments and MySQL for its huge database, which must respond almost in real time to millions of users simultaneously. About the workload of the database, it should be stated that most of the major social networks exploit the system memcached to improve performance of database-driven Web services, saving about specific memories most frequent queries and then speeding the creation of dynamic pages. The database contents is still stored on specific servers, with storage capacities of hundreds of terayte or petabytes.

Open source dominates the sector including on the web server, since Apache is the most popular platform in this context and the same goes for the management of messaging, as a colossus like Twitter using the protocols open source XMPP, also called developed by Jabber and XMPP Standards Foundation to manage the huge amount of messages exchanged daily between users.

From the hardware point of view, the enormous workload is spread across clusters of thousands of servers, most energy efficient, interconnected by specific technologies, the machines are equipped with at least 256 MB of RAM or 512 MB and processor multicore generation. Without such a framework would not be possible to satisfy users and carry out efficient services.

All of these hardware systems are enclosed in specific data centers, which will expand periodically to meet the growing need for space and performance. These infrastructures are a major cost items and are subject to systematic actions to resize the costs, but sometimes result in failures on the service, as occurred last October.