Question

Inherited a database with min size for one data file (primary) at 10 gigs and the min size for a secondary file to 6 gigs.

1. Why would someone have created a secondary data file (NDF)?
2. How do I shrink these below the initial sizes set for the fiels.

Answer 1

1. Why would someone have created a secondary data file (NDF)?

This is a way to get create two files on the operating system. Some advantages:

• 2 file handles
• possibility of splitting across different disks or storage subsystems (fast/slow) even if you don’t do it right away
• prepare for partitioning
• controlling per-file operations, like shrinkfile below
• allows a large db on a prod system to be restored onto a test/dev system that does not have a single large enough disk for single-file. Thus 2x 400GB files on a prod system (same folder) is restored onto 2 different 500GB disks on the dev system.

2. How do I shrink these below the initial sizes set for the files

Use DBCC SHRINKFILE instead of DBCC SHRINKDATABASE. It can be set to any desired size, even smaller than initial, as long as it is greater than the size of data currently stored. To reach smaller sizes, set the target

DBCC SHRINKFILE (1, 0);  -- file id 1, to 0% free space

To find out the file ids, you can use (while in the db)

select * from sysfiles