This link provides a good definition:
“A fault domain is a set of hardware components – computers, switches, and more – that share a single point of failure.” IEEE Computer Magazine March 2011 Issue.
It is a beautiful network architecture concept. Cloud computing applications run on Platform as a Service infrastructures (PaaS) such as Microsoft Azure or Google App Engine. These PaaS applications must be protected against hardware failures such as a disk crash, network fault or server death.
Hence, one of the ways to protect against hardware failures and improve application availability is to setup multiple fault domains for the different instances of an application. As a consequence, when an application temporarily loses some instances, the platform will be able to search the applications topology and point to another fault domain in order to keep the application up and running.