Databases
StackOverflow2013
Postsdbo
POExplain optimal query for large table with clustered index in SQL Server 2008
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POExplain optimal query for large table with clustered index in SQL Server 2008
    primarykey
    data
    text
    <p>I'm working on a very large table (approximately 2.7 million rows added per day), which has the following structure:</p> <pre><code>CREATE TABLE [dbo].[Result]( [ResultDate] [date] NOT NULL, [Thing1Id] [int] NOT NULL, [Num] [int] NOT NULL, [Thing2Id] [int] NOT NULL, CONSTRAINT [PK_Result] PRIMARY KEY CLUSTERED ( [ResultDate] ASC, [Thing1Id] ASC, [Num] ASC )) </code></pre> <p>As the clustered primary key is on ResultDate, Thing1Id and Num, I would expect the following query to be optimal:</p> <pre><code>SELECT Thing2.* FROM dbo.Result INNER JOIN Thing2 ON Thing2.Id = result.Thing2Id WHERE ResultDate &gt;= '2012-01-01' AND ResultDate &lt;= '2012-01-30' AND Thing1Id = 23 </code></pre> <p>As you can see, the query is finding results in Jan-12 for a particular Thing1.</p> <p>However, the execution plan indicates that a huge performance increase can be gained by adding the following index:</p> <pre><code>CREATE NONCLUSTERED INDEX [IX_Missing] ON [dbo].[Result] ([Thing1Id],[ResultDate]) INCLUDE ([Num],[Thing2Id]) </code></pre> <p>And surely enough, adding this index does improve the performance massively.</p> <p>Can someone please explain why? As far as I'm concerned the results should be narrowed sufficiently using the clustered primary key, and adding this will make the index size much larger and add unnecessary overhead.</p> <p>Can I index the table differently to get better performance?</p> <p>(Please note, in reality the table is actually 2 tables unioned, data is shifted from one to the other daily, and the data is partitioned monthly).</p>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. This table or related slice is empty.
    1. USsmartypantsfe
    plurals
    1. PL
      singulars
      1. LTLinked
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    3. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POExplain optimal query for large table with clustered index in SQL Server 2008
      1. This table or related slice is empty.
      1. VTUpMod
    1. COI bet even just adding a non-clustered index on `dbo.Result(Thing2Id)` would speed up your query quite a bit, based on the fact that `Thing2Id` is a foreign key and used in your `INNER JOIN Thing2 ON Thing2.Id = result.Thing2Id` statement...
      singulars
      1. POExplain optimal query for large table with clustered index in SQL Server 2008
      1. USmarc_s
    2. COYes, you're right. I have already added an index on Thing2Id which speeds things up, but I didn't include it in the post because I'm more interested in the clustered index. Thanks.
      singulars
      1. POExplain optimal query for large table with clustered index in SQL Server 2008
      1. USsmartypantsfe
    3. COI don't see why an index on Thing2Id would help for this query. Just adding a index because it has an FK can hurt more than it helps on large tables like this.
      singulars
      1. POExplain optimal query for large table with clustered index in SQL Server 2008
      1. USWim
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload