HOW FACEBOOK HANDLED BILLIONS OF REQUESTS PER SECOND EFFICIENTLY

Summary from the TLDR Web Dev newsletter

Facebook receives billions of requests per second and stores trillions of items in its databases. Traditional web architectures couldn’t keep up with its demands, so the team used a simple key-value store called Memcached. The team scaled Facebook efficiently to become the largest Memcached installation in the world. This article describes how Facebook scaled Memcached across servers, clusters, and regions worldwide in a simple manner.