This is both an excellent post-mortem success story as well as a nice source of scaling wisdom for a distributed system.
http://www.twilio.com/engineering/2014/06/26/how-we-replaced-our-data-pipeline-with-zero-downtime
This is both an excellent post-mortem success story as well as a nice source of scaling wisdom for a distributed system.
http://www.twilio.com/engineering/2014/06/26/how-we-replaced-our-data-pipeline-with-zero-downtime