I have always been of the personal opinion that all SQL Server tables have a clustered index. As part of the research I am doing on a new book on High Performance Index Maintenance, and on heaps specifically, I ran across this SQL Server Best Practices Article from Microsoft.
This article describes a series of tests that Microsoft did using SQL Server 2005 SP1 in 2007. While the article is a little dated, it does a great job of “proving” that every table should have a clustered index. It goes into great detail explaining why using a clustered index instead of a heap is virtually always the best choice.
If you are not already convinced that heaps should be avoided, you will be after reading this paper.
|Share this post :|