About This Blog

.net & SQL Samples, programming tips and tricks, performance tips, guidelines, and best practices

Wednesday, 30 March 2011

How to protect code written in Stored Procedure or User Defined Function?

We have always been wondering on how to protect/secure our t-sql code written in Stored Procedures and UDF in shared hosting environment from other people who might tweak our work. Here, I would discuss a few options I know to serve purpose.
As per my knowledge, MS SQL Server doesn’t help much with it. The possible alternatives which could be tried are –
  1.        Making Use of WITH ENCRYPTION option

The normal way of creating a stored procedure is -

CREATE PROCEDURE dbo.SimpleStoredProc
    SELECT 'Some t-sql statements'

Creating the stored procedure using the WITH EXCRYPTION option is -

CREATE PROCEDURE dbo.EncryptedStoredProc
    SELECT 'Some t-sql statements'

However, please make sure that the code of the SP is backed up as a separate script file for future references.

Now, when we try to use the following command to get the details of the SP -
EXEC sp_helptext 'dbo.EncryptedStoredProc'

We get the following error message –
The object comments have been encrypted.

And when we try to open this encrypted SP using SSMS, we get the following error message –

Microsoft SQL-DMO
Error 20585: [SQL-DMO]
    Encrypted object is not transferable,
    and script can not be generated.

a.       Once encrypted, it is not possible to decrypt using SQL Server commands.

a.       We as developers will always have to keep a copy of the SP/UDF as a script for our reference or future updates.
b.      One way I know using which this approach could be defeated is by running SQL Profiler while executing the stored procedure.
c.       Another way that users might use to get at your encrypted code is by using readily available code (if any) that allows you to break SQL Server's relatively trivial encryption algorithm. 
2.       Stop creating Stored Procedures and Functions

The very first question which comes to our mind is – if we stop creating them, then what is the alternative. We might use an of the below stated technique –
a.       Make use of Parameterized queries directly in our source code. As the code is deployed in the form of an assembly, it becomes hard to directly get into our query logic. However, de-compilation is always possible and to tackle it we can always obfuscate our assemblies. To learn more about obfuscation please refer http://msdn.microsoft.com/en-us/magazine/cc164058.aspx
b.      Secondly, we can make use of sql-clr functions wherever possible. This would again help us in hiding our logic from the preying eyes. Please refer http://msdn.microsoft.com/en-us/library/ms345136(v=sql.90).aspx for more details.

One should not make use of WITH ENCRYPTION option unless it’s the last option and have a thorough knowledge of its consequences after implementation. However, I would strongly recommend NOT TO USE it.

Saturday, 12 March 2011

Table Valued parameters in SQL Server 2008

The most awaited feature has arrived - "Table-Valued parameters in SQL Server 2008". Here, I would discuss on how to use this new feature with .net.

Problem Statement
Before to the introduction of table-valued parameters in SQL Server 2008, the options for passing data from multiple rows of a DataTable to a stored procedure or a parameterized SQL command were limited. Some of the possible options for a developer until now were:
1. Make multiple parameters, with one parameter representing one column of the DataTable
2. Create XML string & then parse it in the SP
3. Create a delimiter separated string & then parse it in the SP

Also, the above stated options had to be repeated for each row....hhuuuuuuuuhh.......just imagine the round trips to the server for large DataTables or the complex logic to be implemented for parsing the parameters.

Solution with Table-Valued parameter
1. Creating Table-Valued Parameter Types
( Col1 int, Col2 nvarchar(50),Coln ... )

2. Create Stored Procedure with a table valued parameter
(@TableType dbo.TableType READONLY)

NOTE : The READONLY keyword is required for declaring a table-valued parameter.

3. Use the parameter as under -
For Insert
INSERT INTO dbo.Tbl (Col1,Col2)
SELECT paramTbl.Col1, paramTbl.Col2
FROM @TableType AS paramTbl;

For Update
UPDATE dbo.Tbl
SET Tbl.Col2 = paramTbl.Col2
FROM dbo.Tbl
INNER JOIN @TableType AS paramTbl
ON dbo.Tbl.Col1 = paramTbl.Col1;

For Delete
FROM dbo.Tbl
INNER JOIN @TableType AS paramTbl
ON dbo.Tbl.Col1 = paramTbl.Col1;

4. Calling from .net Code

// Assumes connection is an open SqlConnection object.
using (connection)
{// Create a DataTable with the modified rows.
DataTable addedValues = ValuesDataTable.GetChanges(DataRowState.Added);
 // Configure the SqlCommand and SqlParameter.
SqlCommand insertCommand = new SqlCommand("SP_Name", connection);
 insertCommand.CommandType = CommandType.StoredProcedure;
 SqlParameter tvpParam = insertCommand.Parameters.AddWithValue("@TableType ", addedValues );
tvpParam.SqlDbType = SqlDbType.Structured;
 // Execute the command.

Limitations of Table-Valued parameter
1. User-defined functions do not support Table valued parameters.
2. Table-valued parameters can only be indexed to support UNIQUE or PRIMARY KEY constraints.
3. SQL Server does not maintain statistics on table-valued parameters.
4. Table-valued parameters are read-only in Transact-SQL code. You cannot update the column values in the rows of a table-valued parameter and you 
cannot insert or delete rows. To modify the data that is passed to a stored procedure or parameterized statement in table-valued parameter, you must 
insert the data into a temporary table or into a table variable.
5. You cannot use ALTER TABLE statements to modify the design of table-valued parameters.
6. If you are using Entity Framework 1.0, DBType enum does not support Structured type.
As a work around, create a SQLParameter & then typecast it into DbParameter.

SqlParameter param = new SqlParameter("@TableType", addedValues );
param.SqlDbType = SqlDbType.Structured;
command.Parameters.Add(param as DbParameter);