A Closer Look at Tumblr's Architecture update from February 2012

High Scalability regularly profiles the scalability challenges of some of the web's busiest sites. Today HS provides a closer look at the architecture behind architecture behind Tumblr, the popular instablogging service that has grown to 15 billion page views per month.

That kind of growth can strain a startup's infrastructure, and Tumblr has had its share of outages, to the point where one of its users created a mascot for the site's downtime error screens.

"Tumblr started as a fairly typical large LAMP application," High Scalability writes. "The direction they are moving in now is towards a distributed services model built around Scala, HBase, Redis, Kafka, Finagle,  and an intriguing cell based architecture for powering their Dashboard. Effort is now going into fixing short term problems in their PHP application, pulling things out, and doing it right using services."

An interesting data point: Tumblr currently runs its entire operation out of a single colocation facility. Read more at High Scalability.