I remember that before I had Nimble in the rack that I would wake up worrying about my backups. Did the FULL to tape complete? Will the tape work when I need it to? Will Iron Mountain get it delivered quick enough?
With Nimble all of those questions were answered. I have CS240's in my production datacenter replicating to my DR/BC site in Las Vegas.
I have 2 arrays in Garden Grove, CA each with three 1TB storage LUNs and one 1TB log LUN assigned to my Active/Passive MS Cluster Exchange 2007 Mailbox server. I have about 2.5TBs of mail over 1500 mailboxes. I've configured a "Volume Collection" with the Exchange LUNs and set a daily snap that replicates to the DR/BC site as well as a weekly "schedule" that creates a FULL with a longer life span than the dailies.
When a restore is needed I clone the appropriate snap, assign it to the servers, mount the LUN and assign an unused drive letter (typically R:\ - recovery). I use Lepide "Exchange Recovery Manager" software to open the offline .edb file, browse and select what I need to restore, specify a destination (I usually export to .pst files which I inject or send to the user).
That's it... no recovery groups... no merge worries... simple!