Using WITH ROWCOUNT and PAGECOUNT option of SQL Server UPDATE STATISTICS

By: Ben Snaidero   |   Read Comments (6)   |   Related Tips: More > Query Plans

Problem

Ever wondered what the SQL Server explain plan for you query would look like if there were millions of records in the tables of your SQL Server database. This is a question that SQL Server database administrators (DBAs) as well as developers often ask and in most cases what they do is load a bunch of dummy data into the tables and check the explain plan that is generated. If there is a lot of referential integrity defined on the SQL Server database this can become a very time consuming task. In other cases you may not have the necessary disk space in your development environment to load massive amounts of data into the tables. This tip will look at an alternative to this method.

Solution

Available since SQL Server 2005 have been the undocumented options of the UPDATE STATISTICS command, WITH ROWCOUNT and WITH PAGECOUNT. Using these options allows you to trick the optimizer into thinking the tables have many more records and pages than they actually do. Let's use a simple example to illustrate how these options work.

Sample table setup

-- Table creation logic
CREATE TABLE testtable ([col1] [int] NOT NULL primary key,
                        [col2] [int] NULL,
                        [col3] [int] NULL,
                        [col4] [varchar](50) NULL);
                        
CREATE NONCLUSTERED INDEX [IX_testtable] ON testtable ([col2] ASC);
 
-- Populate table
DECLARE @val INT
SELECT @val=1
WHILE @val <= 2000
BEGIN  
   INSERT INTO testtable (col1, col2, col3, col4) 
       VALUES (@val,@val % 10,@val,'TEST' + CAST(@val AS VARCHAR))
   SELECT @val=@val+1
END

Now that we have a small table with some data loaded into it, let's take a look at what the explain plan looks like for the following simple query.

SELECT * FROM testtable WHERE col2 = 6;

We can see from the explain plan that this query is going to do an index scan. Now let's update the statistics on both the table and index using the new options. Here is the syntax:

UPDATE STATISTICS testtable WITH ROWCOUNT = 50000000, PAGECOUNT = 500000
UPDATE STATISTICS testtable IX_testtable WITH ROWCOUNT = 50000000, PAGECOUNT = 500000

Before we take a look at the new explain plan let's take a look at the sys.partitions and sys.dm_db_partition_stats DMVs and see what effect the statistics update had on these objects. Note: The output below shows the DMVs before and after the UPDATE STATISTICS command was run. Only the relevant columns are displayed.

We can see from the output above that the row_count and in_row_data_page_count have been updated based on what we supplied to the UPDATE STATISTICS command. Now let's take a look and see if this has had any effect on the explain plan.

This explain plan shows that the optimizer has now chosen a parallel execution plan. Using these undocumented options we can get an idea of how the explain plan for a query may change due to larger data sets without having to load large amounts of data into our tables.

Next Steps

• Further reading on SQL Server statistics:
  • SQL Server auto update and auto create statistics
  • What are the SQL Server WASYS Statistics
• More information on Explain Plans
  • Query execution plans
  • How to read SQL Server graphical execution plans
• Other Performance Tuning tips

About the author

Ben has been a SQL Server and Oracle DBA for over 10 years and focuses on performance tuning.

View all my tips

We Recommend

• More Database Developer Tips...

Print

Tweet

Become a paid author

Comments and Feedback:

Post a Comment or Question

Keep it clean and stay on the subject or we may delete your comment.
Your email address is not published. Required fields are marked with an asterisk (*)

*Name

*Email

Notify for updates Signup for our newsletter

Comments *** NOTE *** - If you want to include code from SQL Server Management Studio (SSMS) in your post, please copy the code from SSMS and paste the code into a text editor like NotePad before copying the code below to remove the SSMS formatting.
*Enter Code