Can you perform a SQL Server filegroup restore using a full backup file as input? I couldn’t find anything that specifically stated that in documentation, so I tried it. YES you can! It worked great! I wanted to use this technique to create a testing database without copying the data from a particular set of tables (in this case it was filestream data) on a different server. I just used the existing full backup file from my production database and ran a RESTORE command. Here are a couple of notes about it:
- The database cannot previously exist, so you must drop the database first.
- You must restore (at a minimum) the primary filegroup and the log, using the WITH REPLACE option.
I cut about 66% of the restore time off (since there were many filestream documents in the filegroup that I skipped).