SQL Server 2012中快速插入批量数据的示例及疑惑

SQL Server 2008中SQL应用系列--目录索引

今天在做一个案例演示时，在SQL Server 2012中使用Insert语句插入1万条数据，结果遇到了一个奇怪的现象，现将过程分享出来，以供有兴趣的同学参考。

附：我的测试环境为：
SQL Server 2012，命名实例

Microsoft SQL Server 2012 - 11.0.2100.60 (Intel X86)
Feb 10 2012 19:13:17
Copyright (c) Microsoft Corporation
Enterprise Edition on Windows NT 6.0 <X86> (Build 6002: Service Pack 2)

创建示例数据库

[sql] view plaincopyprint?

1. IF OBJECT_ID('DemoPager2012')IS NOTNULL
2. DROP DataBase DemoPager2012
3. GO
4. 
   
5. CREATE Database DemoPager2012
6. GO
7. 
   
8. USE DemoPager2012
9. GO

IF OBJECT_ID('DemoPager2012') IS NOT NULLDROP DataBase DemoPager2012GOCREATE Database DemoPager2012GOUSE DemoPager2012GO

示例表，该表只有四个字段。

[sql] view plaincopyprint?

1. /*
2. Setup script to create the sampletable and fill itwith
3. sample data.
4. */
5. IF OBJECT_ID('Customers','U')IS NOTNULL
6. DROP TABLE Customers
7. 
   
8. CREATE TABLE Customers ( CustomerIDINT primarykey identity(1,1),
9. CustomerNumber CHAR(4),
10. CustomerName VARCHAR(50),
11. CustomerCity VARCHAR(20) )
12. GO

/*Setup script to create the sample table and fill it withsample data.*/IF OBJECT_ID('Customers','U') IS NOT NULLDROP TABLE CustomersCREATE TABLE Customers ( CustomerID INT primary key identity(1,1),CustomerNumber CHAR(4),CustomerName VARCHAR(50),CustomerCity VARCHAR(20) )GO

现在展示批量插入10000条数据到该表中，语句如下：

[sql] view plaincopyprint?

1. TRUNCATE table Customers
2. GO
3. 
   
4. ----清除干扰查询
5. DBCC DROPCLEANBUFFERS
6. DBCC FREEPROCCACHE
7. 
   
8. SET STATISTICS IOON;
9. SET STATISTICSTIME ON;
10. GO
11. 
    
12. DECLARE @d Datetime
13. SET @d=getdate();
14. 
    
15. declare @i int=1
16. while @i<=10000
17. begin
18. INSERT INTO Customers (CustomerNumber, CustomerName,
19. CustomerCity)
20. SELECT REPLACE(STR(@i, 4),' ', '0'),'Customer ' + STR(@i,6),
21. CHAR(65 + (@i % 26)) + '-City'
22. set @i=@i+1
23. end
24. 
    
25. select [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())
26. 
    
27. SET STATISTICS IOOFF ;
28. SET STATISTICSTIME OFF;
29. GO

TRUNCATE table CustomersGO----清除干扰查询DBCC DROPCLEANBUFFERSDBCC FREEPROCCACHESET STATISTICS IO ON;SET STATISTICS TIME ON;GODECLARE @d DatetimeSET @d=getdate();declare @i int=1while @i<=10000beginINSERT INTO Customers (CustomerNumber, CustomerName,CustomerCity)SELECT REPLACE(STR(@i, 4), ' ', '0'),'Customer ' + STR(@i,6),CHAR(65 + (@i % 26)) + '-City'set @i=@i+1endselect [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())SET STATISTICS IO OFF ;SET STATISTICS TIME OFF;GO

该插入语句在SQL Server 2008 r2版本和SQL Server 2012版本中，测试结果如下：

令我惊讶的是，SQL Server 2012居然耗时达到5分多钟，而SQL Server 2008R2版，只需要大约6秒钟。更令人费解的是：查询的I/O统计和elapsed time，在这两个版本中几乎一样。对此异象，我只能理解为每次Insert时的毫秒级精度可能不足以度量该次操作带来的细小差距，然而累积起来就非常可观了。

解决方案一：使用 Set NoCount On，效果立竿见影

[sql] view plaincopyprint?

1. TRUNCATE table Customers
2. GO
3. 
   
4. ----清除干扰查询
5. DBCC DROPCLEANBUFFERS
6. DBCC FREEPROCCACHE
7. 
   
8. SET STATISTICS IOON;
9. SET STATISTICSTIME ON;
10. GO
11. 
    
12. DECLARE @d Datetime
13. SET @d=getdate();
14. set nocount on
15. declare @i int=1
16. while @i<=10000
17. begin
18. INSERT INTO Customers (CustomerNumber, CustomerName,
19. CustomerCity)
20. SELECT REPLACE(STR(@i, 4),' ', '0'),'Customer ' + STR(@i,6),
21. CHAR(65 + (@i % 26)) + '-City'
22. set @i=@i+1
23. end
24. 
    
25. select [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())
26. 
    
27. SET STATISTICS IOOFF ;
28. SET STATISTICSTIME OFF;
29. GO

TRUNCATE table CustomersGO----清除干扰查询DBCC DROPCLEANBUFFERSDBCC FREEPROCCACHESET STATISTICS IO ON;SET STATISTICS TIME ON;GODECLARE @d DatetimeSET @d=getdate();set nocount ondeclare @i int=1while @i<=10000beginINSERT INTO Customers (CustomerNumber, CustomerName,CustomerCity)SELECT REPLACE(STR(@i, 4), ' ', '0'),'Customer ' + STR(@i,6),CHAR(65 + (@i % 26)) + '-City'set @i=@i+1endselect [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())SET STATISTICS IO OFF ;SET STATISTICS TIME OFF;GO

Set NoCount On（http://msdn.microsoft.com/zh-cn/library/ms189837.aspx）的作用：使返回的结果中不包含有关受 Transact-SQL 语句影响的行数的信息。这在批量插入时将显著提高性能。至于 本例中，为什么SQL Server 2008 R2版中却不受该开关影响，希望知道的同学不吝赐教，非常感谢。

改进解决方案二：使用 Set NoCount On+Transaction

[sql] view plaincopyprint?

1. TRUNCATE table Customers
2. GO
3. 
   
4. ----清除干扰查询
5. DBCC DROPCLEANBUFFERS
6. DBCC FREEPROCCACHE
7. 
   
8. 
   
9. 
   
10. SET STATISTICS IOON;
11. SET STATISTICSTIME ON;
12. GO
13. 
    
14. DECLARE @d Datetime
15. SET @d=getdate();
16. set nocount on
17. declare @i int=1
18. BEGIN TRANSACTION
19. while @i<=10000
20. begin
21. INSERT INTO Customers (CustomerNumber, CustomerName,
22. CustomerCity)
23. SELECT REPLACE(STR(@i, 4),' ', '0'),'Customer ' + STR(@i,6),
24. CHAR(65 + (@i % 26)) + '-City'
25. set @i=@i+1
26. end
27. COMMIT
28. select [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())
29. 
    
30. SET STATISTICS IOOFF ;
31. SET STATISTICSTIME OFF;
32. GO

TRUNCATE table CustomersGO----清除干扰查询DBCC DROPCLEANBUFFERSDBCC FREEPROCCACHE SET STATISTICS IO ON;SET STATISTICS TIME ON;GODECLARE @d DatetimeSET @d=getdate();set nocount ondeclare @i int=1BEGIN TRANSACTIONwhile @i<=10000beginINSERT INTO Customers (CustomerNumber, CustomerName,CustomerCity)SELECT REPLACE(STR(@i, 4), ' ', '0'),'Customer ' + STR(@i,6),CHAR(65 + (@i % 26)) + '-City'set @i=@i+1endCOMMITselect [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())SET STATISTICS IO OFF ;SET STATISTICS TIME OFF;GO

解决方案三：使用递归CTE插入

[sql] view plaincopyprint?

1. TRUNCATE table Customers
2. GO
3. 
   
4. DBCC DROPCLEANBUFFERS
5. DBCC FREEPROCCACHE
6. 
   
7. SET STATISTICS IOON;
8. SET STATISTICSTIME ON;
9. GO
10. 
    
11. DECLARE @d Datetime
12. SET @d=getdate();
13. 
    
14. /*****运用CTE递归插入，速度较快，邀月注***********************/
15. WITH Seq (num,CustomerNumber, CustomerName, CustomerCity)AS
16. (SELECT 1,cast('0000'asCHAR(4)),cast('Customer 0'AS NVARCHAR(50)),cast('X-City'as NVARCHAR(20))
17. UNION ALL
18. SELECT num + 1,Cast(REPLACE(STR(num, 4),' ', '0')AS CHAR(4)),
19. cast('Customer ' + STR(num,6)AS NVARCHAR(50)),
20. cast(CHAR(65 + (num % 26)) +'-City' AS NVARCHAR(20))
21. FROM Seq
22. WHERE num <= 10000
23. )
24. INSERT INTO Customers (CustomerNumber, CustomerName, CustomerCity)
25. SELECT CustomerNumber, CustomerName, CustomerCity
26. FROM Seq
27. OPTION (MAXRECURSION 0)
28. 
    
29. select [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())
30. 
    
31. SET STATISTICS IOOFF ;
32. SET STATISTICSTIME OFF;
33. GO

TRUNCATE table CustomersGODBCC DROPCLEANBUFFERSDBCC FREEPROCCACHESET STATISTICS IO ON;SET STATISTICS TIME ON;GODECLARE @d DatetimeSET @d=getdate();/*****运用CTE递归插入，速度较快，邀月注***********************/WITH Seq (num,CustomerNumber, CustomerName, CustomerCity) AS(SELECT 1,cast('0000'as CHAR(4)),cast('Customer 0' AS NVARCHAR(50)),cast('X-City' as NVARCHAR(20))UNION ALLSELECT num + 1,Cast(REPLACE(STR(num, 4), ' ', '0') AS CHAR(4)),cast('Customer ' + STR(num,6) AS NVARCHAR(50)),cast(CHAR(65 + (num % 26)) + '-City' AS NVARCHAR(20))FROM SeqWHERE num <= 10000)INSERT INTO Customers (CustomerNumber, CustomerName, CustomerCity)SELECT CustomerNumber, CustomerName, CustomerCityFROM SeqOPTION (MAXRECURSION 0)select [语句执行花费时间(毫秒)]=datediff(ms,@d,getdate())SET STATISTICS IO OFF ;SET STATISTICS TIME OFF;GO

小结：SQL Server 2012中批量插入数据时，请记得Set NoCount ON，并尽可能加上Transaction，当然，推荐使用CTE，这可能会带来性能上的巨大提升。

邀月补充：

后来与微软亚太工程师多次沟通，得出初步结论：

在不打开“set nocount on”时，SSMS 2012与SSMS 2008r2版本的UI在执行效率上可能有极大差异，而与SQL Server引擎没有明显相关。

邀月注：本文版权由邀月和CSDN共同所有，转载请注明出处。
助人等于自助! [e-mail]