When To Define Clustered Index Columns On Non-Clustered Indexes

Kim Tripp explains when to include a clustered index column on a non-clustered index column’s definition:

So, when SHOULD you explictly define the clustering key columns in a nonclustered index? When they ARE needed by the query.

This sounds rather simple but if the column is USED by the query then the index MUST have the column explicitly defined. Yes, I realize that SQL Server will add it… so it’s not necessary NOW but what if things change? (this is the main point!)

One of the more common cases I could think of is multi-part clustered indexes, like on a junction table.

Related Posts

Making Non-SARGable Queries SARGable with an Index

Denis Gobo violates Betteridge’s Law of Headlines: This question came up the other day from a co-worker, he said he couldn’t change a query but was there a way of making the same query produce a better plan by doing something else perhaps (magic?) He said his query had a WHERE clause that looked like […]

Read More

A Sizing Problem with Heaps

Slava Murygin shows an example where adding a clustered index to an existing table can reduce its size: Tables were very narrow with just a few columns and my expectations for data growth were very modest. However, after just a little while I was very surprised when my database showed huge unexpected growth and size […]

Read More

Categories

February 2017
MTWTFSS
« Jan Mar »
 12345
6789101112
13141516171819
20212223242526
2728