Passing variables into Openquery and SQL Injection - sql-server

I have two databases (A and B), both SQL Server, on different servers. These databases are connected with a linked server.
I have to be able to insert rows with distinct values into a table in database B using a stored procedure on database A. This stored procedure uses OPENQUERY in order to do the INSERT statements into database B.
I know OPENQUERY does not accept variables for its arguments. OPENQUERY has specific syntax on how to do an insert into a linked DB:
INSERT OPENQUERY (OracleSvr, 'SELECT name FROM joe.titles')
VALUES ('NewTitle');
Nevertheless, the MS documentation shows a way to pass variables into a linked server query like this:
DECLARE #TSQL varchar(8000), #VAR char(2)
SELECT #VAR = 'CA'
SELECT #TSQL = 'SELECT * FROM OPENQUERY(MyLinkedServer,''SELECT * FROM pubs.dbo.authors WHERE state = ''''' + #VAR + ''''''')'
EXEC (#TSQL)
And here is the issue. Lets say the table in database B has two columns, ID (int) and VALUE (nvarchar(max))
Thus, for a stored procedure to be able to insert different values into a table in database B, my procedure looks like this:
CREATE PROCEDURE openquery_insert
#var1 int,
#var2 nvarchar(max)
AS
BEGIN
SET NOCOUNT ON;
BEGIN
DECLARE #SQL_string nvarchar(max)
SET #SQL_string = 'insert openquery(LINKEDSERVER, ''SELECT ID, VALUE from TABLE'') VALUES ('
+ CAST(#var1 AS NVARCHAR(5)) + ', '
+ '''' + CAST(#var2 AS NVARCHAR(max)) + ''''
+ ')'
EXEC sp_executesql #SQL_string
END
END
The procedure can be called as
EXEC openquery_insert #var1 = 1, #var2 = 'asdf'
But if #var2 were to be ' DROP TABLE B--, a SQL injection attack would be successful.
Is there a way in order to prevent SQL Injection with OPENQUERY?
I do not control what the values are for the arguments #var1 and #var2 when the procedure gets called
I am not able to create functions or stored procedures on database B
I have to use OPENQUERY, I can not use four part naming in order to do the insert
I have to use a stored procedure on DB A
Thanks!

The "hacky" way is to insert your arguments into a local table first and then do the INSERT INTO ... SELECT using OPENQUERY.
This is all straightforward if your SP is ever called by one process in a synchronous fashion: you can have one table where you insert values into then execute OPENQUERY to grab them and then you delete those values from the table.
If concurrency is a requirement then you have to write logic that creates uniquely named tables etc. which quickly becomes somewhat messy.

Related

How to query from the dynamic database name inside the function

ALTER FUNCTION FnVersion
(#DBName NVARCHAR(255),
#ID INT)
RETURNS #TABLE TABLE (iD INT, VersionNo INT)
AS
BEGIN
DECLARE #SQL VARCHAR(2000)
SET #SQL = #DBName
SELECT #SQL = 'SELECT iD, VersionNo FROM' + #DBName + '.dbo.ConfigInfo WHERE IdValue = #ID'
EXECUTE SQL
RETURN;
END
The db name will be passed as input and the query will be done on dynamic database
The above logic of exec SQL statement caused an error.
What is the way to achieve dynamic query from dynamic database?
You can not use dynamic SQL in SQL Function. but make it part of procedure and store those data into temp table with remote procedure call. Performance will be better to use remote procedure rather than remote query.
you can use INSERT INTO with EXEC command. and manipulate temp table with your logic. it would be better approach.

Stored Procedure in sql for selecting columns based on input values

I am trying to code a stored procedure in SQL that does the following
Takes 2 inputs (BatchType and "Column Name").
Searches database and gives the batchdate and the data in the column = "Column name"
Code is as give below
ALTER PROCEDURE [dbo].[chartmilldata]
-- Add the parameters for the stored procedure here
(#BatchType nvarchar (50),
#Data nvarchar(50))
AS
BEGIN
-- Insert statements for procedure here
SELECT BatchDate,#Data FROM --Database-- WHERE BatchType = #BatchType
END
I am trying to select column from the database based on operator input. But I am not getting the output. It would be great if someone can give me a direction.
You may want to build out your SELECT statement as a string then execute it using sp_executesql.
See this page for more info:
https://msdn.microsoft.com/en-us/library/ms188001.aspx
This will allow you to set your query to substitute in your column name via your variable and then execute the statement. Be sure to sanitize your inputs though!
You'd need to use dynamic SQL, HOWEVER I would not recommend this solution, I don't think there is anything I can add as to why I wouldn't recommend it that isn't explained better in Erland Sommarskog in The Curse and Blessings of Dynamic SQL.
Nonetheless, if you had to do it in a stored procedure you could use something like:
ALTER PROCEDURE [dbo].[chartmilldata]
-- Add the parameters for the stored procedure here
(#BatchType nvarchar (50),
#Data nvarchar(50))
AS
BEGIN
-- DECLARE AND SET SQL TO EXECUTE
DECLARE #SQL NVARCHAR(MAX) = N'SELECT BatchDate = NULL, ' +
QUOTENAME(#Data) + N' = NULL;';
-- CHECK COLUMN IS VALID IN THE TABLE
IF EXISTS
( SELECT 1
FROM sys.columns
WHERE name = #Data
AND object_id = OBJECT_ID('dbo.YourTable', 'U')
)
BEGIN
SET #SQL = 'SELECT BatchDate, ' + QUOTENAME(#Data) +
' FROM dbo.YourTable WHERE BatchType = #BatchType;';
END
EXECUTE sp_executesql #SQL, N'#BatchType NVARCHAR(50)', #BatchType;
END
It would probably be advisable to change your input parameter #Data to be NVARCHAR(128) (or the alias SYSNAME) though, since this is the maximum for column names.

How to check existence of a table from a different sql db?

I have db A and db B. At the beginning of a stored procedure I want to back up all rows from B.mytable to B.mytablebackup. The rest of the stored procedure runs against tables on db A (which gathers data and writes it to B.mytable).
So I check to see if B.mytablebackup exists
IF EXISTS(SELECT 1 FROM B.dbo.mytablebackup)
and if it does, the stored procedure does an
INSERT INTO B..mytablebackup SELECT * FROM B..mytable
If it doesn't exist it does a
SELECT * INTO B..mytablebackup from B..mytable
But when I execute the stored procedure I get the error
There is already an object named 'mytablebackup' in the database
I added a Print statement and execution is taking the "does not exist" branch of the IF.
What am I doing wrong?
For SQL Server, you should use system view sys.tables to check if table exists.
IF EXISTS(SELECT 1 FROM B.sys.tables WHERE name = 'mytablebackup')
OBJECT_ID can be used too:
IF OBJECT_ID('B.dbo.mytablebackup') IS NOT NULL
You can directly check from the given DB,SCHEMA and TABLE parameters (For dynamic database, schema and table use)
DECLARE #targetdatabase NVARCHAR(MAX),
#SchemaName NVARCHAR(MAX),
#TableName NVARCHAR(MAX)
DECLARE #TempTableName NVARCHAR(MAX) = QUOTENAME(#targetdatabase) + '.' +
QUOTENAME(#SchemaName) + '.' + QUOTENAME(#TableName)
IF OBJECT_ID(#TempTableName) IS NULL
BEGIN
PRINT #TempTableName
END

Replace first FROM in sql query

I need to write a query engine on a web app, what needs to be accomplish is that a user can enter any SELECT statement into a textbox and then the results should be created into a new table.
This is my function I have created but it only support SQL Server 2012 and I want similar to this function but only it should support SQL Server 2005 and above:
CREATE FUNCTION [dbo].[CustomQueryTableCreation]
(
#TableName varchar(max),
#sql NVARCHAR(MAX)
)
RETURNS
#TableBuilder TABLE
(
DS varchar(max)
)
BEGIN
INSERT INTO #TableBuilder
SELECT 'CREATE TABLE dbo.' + #TableName+'(';
INSERT INTO #TableBuilder
SELECT
CASE column_ordinal
WHEN 1 THEN '' ELSE ',' END
+ name + ' ' + system_type_name + CASE is_nullable
WHEN 0 THEN ' not null' ELSE '' END
FROM
sys.dm_exec_describe_first_result_set
(
#sql, NULL, 0
) AS f
ORDER BY
column_ordinal;
INSERT INTO #TableBuilder
SELECT ');';
RETURN
END
What I want to do now is that I want to search through my query and replace the FIRST FROM with INTO NewTable FROM.
The query can contain multiple joins.
Should I control this with SQL or C#?
I had a similar problem with the 2005 Environment. If you save the Select query to a table, and use the following built in procedure to execute the query:
EXECUTE sp_executesql #Query
Here is the MS docs:
http://msdn.microsoft.com/en-us/library/ms188001%28v=sql.90%29.aspx
Edit
Keeping this in mind, can take the SQL dumps and Create OpenRowset Queries to take the SQL and dump them into a TempTable, and from the Temp Table to a permanent table if required.
I created the following SP's to assist with getting the info to a permanent table.
First the procedure to execute the specific SQL Statement
CREATE PROCEDURE [dbo].[spExecuteRowset]
(
#Query NVARCHAR(MAX)
)
AS
BEGIN
--Execute SQL Statement
EXECUTE sp_executesql #Query
END
Then the OpenRowset SP:
CREATE PROCEDURE [dbo].[spCustomquery]
(
#ProQuery NVARCHAR(MAX),
#Tablename NVARCHAR(MAX)
)
AS
BEGIN
--Insert the info into a Specidied Table
DECLARE #Query NVARCHAR(max)
SET #Query = 'SELECT * INTO #MyTempTable FROM OPENROWSET(''SQLNCLI'', ''Server=localhost;Trusted_Connection=yes;'','' EXEC [YOUR DATABASE].dbo.spExecuteRowset' +''''+#ProQuery+''''') SELECT * INTO '+ #Tablename +' FROM #MyTempTable'
--FOR DEBUG ONLY!!!!
PRINT #Query
EXEC [YourDatabase].dbo.spExecuteRowset #Query
END
This takes it from the #tempTable to A Physical Table.
Here are some docs on OpenRowset.
You have no guarantee that the first from in a query will accept an into, because you can have a subselect in the select statement. In addition, you could have a field name like datefrom that throws things off too.
But, assuming you have "simple" SQL statements, you can do it as:
select stuff(#query, charindex('from ', #query), 0, 'into '+#Table+' ')
from t;
EDIT:
The following is what you really want to do:
select *
into #Table
from (#query) q;
Using the subquery solves the problem.
This is a well-known problem. String concatenation is usually a bad/limited solution.
The more recommended solution is to let some other mechanism to return you the result set (openquery etc.), and then insert it to a table.
For example:
SELECT *
INTO YourTable
FROM OPENQUERY([LinkedServer],your query...)

Dynamically named temp table returns "invalid object name" when referenced in stored procedure

When I run the following code, I get an "invalid object name" error, any idea why?
I need to create a dynamically named temp table to be used in a stored procedure.
DECLARE #SQL NVARCHAR(MAX)
DECLARE #SessionID NVARCHAR(50)
SET #SessionID = 'tmp5l7g9q3l1h1n5s4k9k7e'
;
SET
#SQL = N' CREATE TABLE #' + #SessionID + ' ' +
N' (' +
N' CustomerNo NVARCHAR(5), ' +
N' Product NVARCHAR(3), ' +
N' Gross DECIMAL(18,8) ' +
N' )'
;
EXECUTE sp_executesql #SQL
;
SET
#SQL = N' SELECT * FROM #' + #SessionID
;
EXECUTE sp_executesql #SQL
Thanks!
WHY MESS WITH THE NAMES? Let SQL Server will manage this for you:
Temporary Tables in SQL Server
from the above link:
If the same routine is executed simultaneously by several processes,
the Database Engine needs to be able to distinguish between the
identically-named local temporary tables created by the different
processes. It does this by adding a numeric string to each local
temporary table name left-padded by underscore characters. Although
you specify the short name such as #MyTempTable, what is actually
stored in TempDB is made up of the table name specified in the CREATE
TABLE statement and the suffix. Because of this suffix, local
temporary table names must be 116 characters or less.
If you’re interested in seeing what is going on, you can view the
tables in TempDB just the same way you would any other table. You can
even use sp_help work on temporary tables only if you invoke them from
TempDB.
USE TempDB
go
execute sp_Help #mytemp
or you can find them in the system views of TempDB without swithching
databases.
SELECT name, create_date FROM TempDB.sys.tables WHERE name LIKE '#%'
You are doing it wrong!
Try:
exec(#SQL)
instead of:
EXECUTE sp_executesql #SQL
To use sp_executesql the variable must be inside #SessionID the quotes and it must be provided has input parameter. Check this for a full example!
You've to be aware that Dynamic SQL is a good port for SQL injections!
This syntax works
CREATE TABLE #SessionID (CustomerNo NVARCHAR(5), Product NVARCHAR(3), Gross DECIMAL(18,8));
Select COUNT(*) from #SessionID;
Drop Table #SessionID;

Resources