Facebook operates at massive scale and is globally interconnected. Rendering a single page on Facebook involves hundreds of machines examining tens of thousands of pieces of data from dozens of services - all in real time.
To meet this challenge, Facebook’s infrastructure team rethought every layer of the technology stack. Our development cycle is extremely fast, and we've built tools to keep it that way. It's common to write code and have it running on the live site a few days later.
- Facebook is one of the the largest users in the world of memcached, an open source caching system, and we have one of the largest MySQL database clusters anywhere.
- We store more than 100 petabytes (100 quadrillion bytes) of photos and videos. We’ve built storage and serving technologies, such as Haystack, which allow us to efficiently serve and store data.
- The engineering team developed Hip Hop for PHP - a way to transform PHP source code into highly optimized C++ code. HipHop offers significant performance gains when compared to traditional PHP.
- Apache Hive is our data warehouse infrastructure built on top of Hadoop, which provides tools to enable easy data summarization, adhoc querying and analysis of large datasets.
Open Compute Project
To support the massive scale and rapid growth of our computing needs, we custom designed and built our software, servers and data centers from the ground up. The result: a data center in Prineville, Oregon, which yielded a significant increase in energy efficiency while significantly reducing our costs compared with traditional servers and leased data centers.
We then open-sourced our server and data center designs to help start the Open Compute Project, an industry initiative dedicated to increasing the pace of innovation in data center technology. The Open Compute Project aims to make highly efficient scale computing technology available to everyone, as well as helping reduce the environmental impact of computing infrastructure. See: http://opencompute.org/
Forest City, North Carolina