Yes, I believe it's normal. This is why GRT restores are recommended to be done from disk. If restoring from tape the entire backup must be staged first. Of course, this relies on you having enough free disk space to keep the backups lying around...

I'm sure there's an article on this in the BE Knowledgebase....

Paulo,

 

As stated, this is expected behavior and recommendation is to run to B2D (not counting the current issue with Duplicate to tape).

 

The actual reason for this is that a GRT backup to a B2D folder will keep the exchange database in a dynamic form, so it is read on the fly to do a restore.  This of course cannot be done to a tape device, so when doing a restore, the data is staged.  When it is staged, the data is changed from the tape format to the dynamic database format that the IMG folders use in B2D.  The actual restore takes a few seconds, the staging of the data though can take awhile depending on backup size.

 

Consider this a temporary issue though. Once the hot fix comes out for the duplicate to tape issue then you will be able to change to running the job to B2D and dupicating it to tape.  All your restores can then be done from B2D and thus eliminate the long staging process.