Introduction

One of my friends asked: "Is it better to create an Index within a Stored Procedure?" While it depends on the columns used in a stored procedure to improve the Join performance and Drop the stored procedure after getting the desired output and just before the ending...

Technical Answer

...yes, we can create an Index within the Stored Procedure.

Now, We Have to Think What Our Answer means

Technically, what my friend mentioned in the scenario is possible. But performance wise, it’s again a problematic solution. That’s means to solve something we again create another problem. To understand it properly let’s take an example of such kind of Stored Procedure.

 IF OBJECT_ID(N'[dbo].[sproc_StudentDetails]', N'P')IS NOT NULL
   DROP PROCEDURE [dbo].[sproc_StudentDetails];
GO

CREATE PROCEDURE [dbo].[sproc_StudentDetails]
     (
         @p_StdClass   INT
      )
AS
BEGIN
      -- Creating Non Clustered Index on IDNO
      CREATE NONCLUSTERED INDEX IX_NC_IDNO_StdDtl
              ON tbl_StudentDetails(IDNO);

       CREATE NONCLUSTERED INDEX IX_NC_IDNO_StdMarks
              ON tbl_StudentMarks(IDNO);

      -- Making JOIN on IDNO for Performance Reason
       SELECT a.Roll, a.StdName, b.Marks1, b.Marks2
       FROM   tbl_StudentDetails AS a
              INNER JOIN tbl_StudentMarks AS b ON a.IDNO = b.IDNO;

      -- Droping the Index
       DROP INDEX IX_NC_IDNO_StdDtl ON tbl_StudentDetails;
       DROP INDEX IX_NC_IDNO_StdMarks ON tbl_StudentMarks;
END
GO

Here in this example if we look carefully the No clustered Index is created and after successful joining it again drops.

It is technically perfect.

So, What Is the Problem...

 SELECT a.Roll, a.StdName, b.Marks1, b.Marks2
       FROM   tbl_StudentDetails AS a
              INNER JOIN tbl_StudentMarks AS b ON a.IDNO = b.IDNO;

The performance of the JOIN condition is increased due to a non clustered index. So we think that we got the performance gain...but, in fact, we did NOT.  Always understand fully before implementing.

When the Index created the Index table and the Statistical table, both are updated, so making an index within a stored procedure again takes a long time to create.

By the index creation we solve the Join performance, but Index creation time is again a pain point and slows down the performance of a stored procedure.

Hope you like it.