PowerShell

In this post we are going to show the implementation of a PowerShell script using SMO to move data between filegroups on SQL Server. This article is the second of our two part series on “Moving SQL Server data between filegroups – Database Structures”, you can find the first article here.

The Challenge

Looking around on the web, I couldn’t find a solution to the problem of moving data between filegroups that I liked. Further, many of those solutions are T-SQL based, which I thought were very complex. So I went off to write it myself. The problem lends itself to an iterative solution and I felt that T-SQL was not the right tool for the job. Enter PowerShell, which give us the ability to easily iterate over sets of data with minimal code, couple that with the SQL Server Management Object model and we have the makings of an elegant solution.

Why is moving data between filegroups hard?

****As a consultant its common to walk into a customer site and find databases that are contained in one very large file. For various reasons it can be beneficial to adjust the number a data files for a database. See here. However, in SQL Server moving data from a one file database into a multi-file configuration is a non-trivial task. It’s a two step process, requiring that you add a new filegroup then in the filegroup add your multi-file configuration. Once you have that up, then we need to rebuild the indexes into that filegroup. This can be challenging if you have a lot of tables with a lot of indexes as SSMS allows you do move data but only for non-clustered indexes and only one at a time. Another issue is there are different techniques for moving different physical structures such as clustered indexes, heap and tables with LOB data.