SQL：限制在SQL Server中可以查询的行数

Question

I would like to force limit the number of records being queried from the SQL Server in a way that I can report back the to user that they need to optimize their query.

Intending to use it a reporting tool where let's say the user is not supposed to extract more than 10,000 records.

SET ROWCOUNT 10000 [THROW EXCEPTION IF MORE IS SELECTED]
//user's query
SELECT ....

Is this possible to do?

Answer 1

您可以做的是重写SQL，以将TOP 10000添加到工具生成的任何SQL查询中。

Answer 2

I think you are asking if you can set something global that would cause an error if the result set contained more than a certain number of records.

I also think the answer to that is, "No."

However, depending on how you are constructing and executing the queries you could possible prepend SET ROWCOUNT 10001 -- note the 1 to prevent the runaway queries you want to avoid.

Then append something like @Felipe suggested:

If @@ROWCOUNT = 10001 RAISERROR ('Too many results. Please, optimize your query', 1, 1);

If you are running all queries through some sort of central processor that would be fairly easy.

EDIT:

This should demo the idea in SSMS:

SET ROWCOUNT 2
SELECT 1 UNION SELECT 2
If @@ROWCOUNT = 2 RAISERROR ('Too many results. Please, optimize your query', 1, 1)

If you are not getting an error then it is being masked by something in you code.

Answer 3

In your application you can create a variable that will be populated by the users say @Nrow if they select any value less than 1000 then select their value otherwise just change the value in your code to select TOP 1000 values, something like this..

DECLARE @Nrow INT = 1000;               --<-- Users Value

SET @Nrow = CASE WHEN @Nrow > 1000      --<-- Your check
                 THEN 1000 
            ELSE @Nrow END

SELECT TOP (@Nrow) *                    --<-- Query
FROM dbo.TableName

Answer 4

I'm not sure I understand your question. If the user will simply provide the number of rows he wants, then John Nolan's answer is enough. But, if your client will provide his own query, then maybe @@ROWCOUNT will help you:

*Your client's query goes here

If @@ROWCOUNT > 10000
    BEGIN
    RAISERROR ('Please, optimize your query', 1, 1);
    END

Answer 5

No, actually it isn't. All solutions with TOP and ROWCOUNT will simply limit the output after a query is already done executing, which I believe was what you wanted to avoid in the first place.

What you can is ask SQL Server to just produce an execution plan, and then check the cardinality estimates. However, keep in mind that estimates are just that, the estimates, and could be off.

Answer 6

This uses dynamic SQL, so I don't know if that's a problem. It is set up to stop counting rows if the maximum is exceeded. Also, I created a demo table.

--Create a table for test data (modeled after the Northwind Employees table).
CREATE TABLE EMPLOYEES
(
    EMPLOYEE_ID INT NOT NULL,
    LASTNAME VARCHAR(20) NOT NULL,
    FIRSTNAME VARCHAR(10) NOT NULL,
    TITLE VARCHAR(30),
    TITLE_OF_COURTESY VARCHAR(25),
    BIRTHDATE DATE,
    HIREDATE DATE,
    ADDRESS VARCHAR(60),
    CITY VARCHAR(15),
    REGION VARCHAR(15),
    POSTAL_CODE VARCHAR(10),
    COUNTRY VARCHAR(15),
    HOME_PHONE VARCHAR(24),
    EXTENSION VARCHAR(4),
    PHOTO VARCHAR(255),
    NOTES VARCHAR(2000),
    REPORTS_TO INT,
    CONSTRAINT PK_EMPLOYEES PRIMARY KEY (EMPLOYEE_ID)
);
GO
--end Create a table for test data (modeled after the Northwind Employees table).

--Insert some data
DECLARE @INSERT_DATA_ROW_COUNTER INT
SET @INSERT_DATA_ROW_COUNTER=0
SET NOCOUNT ON
WHILE @INSERT_DATA_ROW_COUNTER<1500
BEGIN
    INSERT INTO EMPLOYEES 
    (EMPLOYEE_ID, 
    LASTNAME,
    FIRSTNAME,
    TITLE,
    TITLE_OF_COURTESY,
    BIRTHDATE,
    HIREDATE,
    ADDRESS,
    CITY,
    REGION,
    POSTAL_CODE,
    COUNTRY,
    HOME_PHONE,
    EXTENSION,
    PHOTO,
    NOTES,
    REPORTS_TO)
    VALUES 
    (@INSERT_DATA_ROW_COUNTER + 1,
    'LAST' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'FIRST' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'TITLE' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'TITLE_OF_COURTESY' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    DATEADD(YEAR, -25, GETDATE()),
    DATEADD(YEAR, -21, GETDATE()),
    'ADDRESS' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'CITY' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'REGION' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    @INSERT_DATA_ROW_COUNTER + 100001,
    'COUNTRY' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 5551230001),
    0,
    'PHOTO' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    'NOTES' + CONVERT(VARCHAR(32), @INSERT_DATA_ROW_COUNTER + 1),
    1)
    SET @INSERT_DATA_ROW_COUNTER=@INSERT_DATA_ROW_COUNTER+1
END
SET NOCOUNT OFF
--end Insert some data

--This is your WHERE CLAUSE for the count and for the data retrieval
DECLARE @WHERE_CLAUSE VARCHAR(MAX)
SET @WHERE_CLAUSE='LASTNAME LIKE ''%0%'''

--This is the maximum number of rows that you would like to allow
DECLARE @MAX_ROW_COUNT int
SET @MAX_ROW_COUNT=400

DECLARE @COUNTROWS int
SET @COUNTROWS=-1
DECLARE @COUNT_SQL NVARCHAR(MAX)
SET @COUNT_SQL='select @COUNTROWS=count(*) from EMPLOYEES WHERE EMPLOYEE_ID IN (SELECT TOP ' + CONVERT(VARCHAR(64), @MAX_ROW_COUNT + 1) + ' EMPLOYEE_ID FROM EMPLOYEES WHERE ' + @WHERE_CLAUSE +')'
exec sp_executesql @COUNT_SQL, N'@COUNTROWS int output', @COUNTROWS output;
IF @COUNTROWS <= @MAX_ROW_COUNT
BEGIN
    DECLARE @SQL VARCHAR(MAX)
    SET @SQL='SELECT * FROM EMPLOYEES WHERE ' + @WHERE_CLAUSE
    EXEC (@SQL)
END
ELSE
BEGIN
    DECLARE @msg VARCHAR(255)
    SET @msg='The limit for this query is ' + CONVERT(VARCHAR(32), @MAX_ROW_COUNT) + ' rows. Change your query as required.'
    RAISERROR (@msg, 1, 1)
END
DROP TABLE EMPLOYEES
GO

Answer 7

select column_name from table_name limit no_of_rows;

it works in MySQL.....not sure about SQL server.