AlwaysOn Availability Groups and AlwaysOn Failure cluster Instances are new features, introduced in SQL Server 2012, that promise the highest level of availability and disaster recovery.
AlwaysOn Availability Groups provides an alternative to database mirroring, and it gives organizations the ability to automatically or manually fail over a group of databases as a single unit, with support for up to four secondaries. The solution provides zero-data-loss protection and is flexible. It can be deployed on local storage or shared storage, and it supports both synchronous and asynchronous data movement. The application failover is very fast, it supports an automatic page repair, and the secondary replicas can be leveraged to offload reporting and a number of maintenance tasks, such as backups.
Watch out a small video of 10 minutes that explains in details about creation of AlwaysOn Availability Groups. The video is prepared by a DBA who wants himself to be called as askadba