When it comes to backing up Exchange Server data, there are a few different things that can complicate the process. One is granularity. It is rare for an administrator to need to recover an entire mailbox database. More often the administrator will need to instead recover a specific mailbox or possibly even an object within the mailbox.
Granular recovery has never been one of Exchange Server's strong points. For example, some versions of Exchange Server require the creation of a recovery storage group or a recovery database prior to performing a restore operation. Once this object has been created, the recovery operation occurs at the database level (which could mean that you have two parallel copies of the same database). Data must then be copied from the newly recovered database to the production database.
Fortunately, many backup applications streamline this process. So, if you are struggling with granular restores of Exchange using native tools, it is probably time to invest in a backup software product that that offers granular restore for Exchange mailboxes.
Another challenge with protecting Exchange Server data is that an Exchange Server organization can span multiple datacenters. A database availability group can be stretched to multiple data centers. Similarly, a hybrid Exchange Server deployment makes use of on premise Exchange Servers and Office 365.
Stretched database availability groups can be challenging to protect because databases often reside in remote data centers. However, Exchange 2013 makes it possible to back up passive database copies. This means that a mailbox database residing in a remote data center can be replicated to the local data center and then the replica (known as a passive database copy) can subsequently be backed up.
Exchange Server data backup and recovery
SharePoint 2013 e-discovery improves Exchange Server data mining