Use T-SQL to Pull a Random Row from a SQL Database

I would like to use a single, random row of data for each execution of a test that is data bound to a SQL Database.

Solution

This is possible with T-SQL. Here's how to do it:

1. Add a SQL Database Source to the project. Here's what the sample database looks like in SQL Server Management Studio Express:

   

2. Bind the Test to the SQL Database. Here's what the Setup Binding dialog looks like before T-SQL is enabled:

   

3. Check Use T-SQL. Enter the following code into the T-SQL Editor section and click Update:

   • SQL

       SELECT TOP 1 [Name], [City], [Email], [Message]
   FROM [myDataBase].[dbo].[table]
   Where Email Like '%domain%'
   ORDER By NEWID()
   
   
   

   </code></pre></div> * Line 1 indicates how many rows and which columns to use.
   
   * Line 2 indicates the database and table name.
   
   * Line 3 indicates whether to filter based on text matching criteria. In this case, since all entries in the Email column contain domain, all rows are returned.
   
   * Line 4 sorts the rows by a unique identifier, which essentially randomizes them in the database.
   
   
   4.&nbsp; Click **Update again** to see the randomization in action. 
   
   ![Random row][3]
   
   5.&nbsp; The following test navigates to a site containing a sample contact form. Note that the Enter text steps already have the corresponding data columns <a href="/features/data-driven-testing/attach-columns-input-values" target="_blank">Attached to Input Values</a>.
   
   
   ![Bind columns][4]
   
   6.&nbsp; Run the test. A random row will be used to data drive the test each time you execute it.
   
   ![Test execution][5]