everything - absolutely everything.

Besides the fact that they are both part of High Availability - they are night and day different.

RAC provides for high availability (and scalability) "in a room". You have two or more computers sharing a common set of disk. All of the computers have read write access to the data in the single (one) database. Users are logged into the database using every computer that you have hooked together into this cluster. If one computer (node) fails in the cluster - work continues as before - the database never goes down. You have less CPU's perhaps during this time, but the database continues running. This protects from system failures, hardware failures, and many software failures (instance failure, OS panic, whatever).


Data Guard however provides for continuity of operations - if the room in which your RAC cluster resides "goes away" (fire, flood, explosion, whatever) data guard is "somewhere else, ready to take over". Data Guard is primarily about failureover in the event of a catastrophic loss of the primary site.

Data Guard provides other features as well - such as the ability to run in a delayed apply mode - so if you "accidently" truncate a table in production (human error) - you can get the data back from your data guard database, instead of backups. You can offload your backups to the data guard site, you can use the data guard site for readonly reporting if you want. But primarily - data guard is about having a fail over site sitting there read to go in the event of something really bad happening to the primary site.