DAG(Database Availability Group) is a new feature provided in Exchange 2010 to facilitate automatic database level recovery. In Exchange 2003/2007, there is no option to recover a single damaged/corrupted mailbox store alone than restoring the DB from backup. In this case, I will loose all the data added to the store after backup. This feature(DAG) of Exchange 2010 avoids this data loss and provides a seem less and automated recovery with up-to-date information regardless of when you last backed up the store.
Few points I have to highlight about DAG
- A DAG can consists of 16 exchange 2010 servers max. Which means that any database in these 16 exchange servers can be failed over to another server in the same DAG incase of curruption. The failover time is as low as 30 seconds and transparent to users.
- DAGs use continuous replication and a subset of Windows Failover Clustering technologies to provide continuous mailbox availability. So you need to install this manually on Exchange 2010 server or it will get installed automatically when you add it to DAG
- It’s needs a file share witness share to maintain the availability of Exchange servers in DAG
- DAG configuration has some similarities with Cluster configuration like, CNO creation, multiple network cards for exchange hear beat data and public data, common IP to access DAG CNO, etc
MsexchangeTeam (it has videos; don’t miss them)