Sql server 优化存储过程

我正在使用SQLServer2008R2。我有这样的存储过程：

ALTER PROCEDURE [dbo].[GetZone]
    -- Add the parameters for the stored procedure here
@TheInsee int
AS
BEGIN
declare @Zone int

declare @CountA int
declare @CountB int
declare @CountC int
declare @CountD int
declare @CountN int
declare @CountP int
declare @CountN1 int
declare @CountN2 int
declare @CountN3 int
declare @CountN4 int
declare @CountN5 int
declare @CountN6 int

SELECT @CountA = COUNT(*) FROM LS_CLIENT_INSEE_A WHERE NO_INSEE = @TheInsee
SELECT @CountB = COUNT(*) FROM LS_CLIENT_INSEE_B WHERE NO_INSEE = @TheInsee
SELECT @CountC = COUNT(*) FROM LS_CLIENT_INSEE_C WHERE NO_INSEE = @TheInsee
SELECT @CountD = COUNT(*) FROM LS_CLIENT_INSEE_D WHERE NO_INSEE = @TheInsee
SELECT @CountN = COUNT(*) FROM LS_CLIENT_INSEE_N WHERE NO_INSEE = @TheInsee
SELECT @CountP = COUNT(*) FROM LS_CLIENT_INSEE_P WHERE NO_INSEE = @TheInsee

SELECT @CountN1 = COUNT(*) FROM LS_CLIENT_INSEE_N1 WHERE NO_INSEE = @TheInsee
SELECT @CountN2 = COUNT(*) FROM LS_CLIENT_INSEE_N2 WHERE NO_INSEE = @TheInsee
SELECT @CountN3 = COUNT(*) FROM LS_CLIENT_INSEE_N3 WHERE NO_INSEE = @TheInsee
SELECT @CountN4 = COUNT(*) FROM LS_CLIENT_INSEE_N4 WHERE NO_INSEE = @TheInsee
SELECT @CountN5 = COUNT(*) FROM LS_CLIENT_INSEE_N5 WHERE NO_INSEE = @TheInsee
SELECT @CountN6 = COUNT(*) FROM LS_CLIENT_INSEE_N6 WHERE NO_INSEE = @TheInsee

set @Zone = 
      CASE 
         WHEN @CountA >  0   THEN 1
         WHEN @CountB >  0   THEN 2
         WHEN @CountC >  0   THEN 3   
         WHEN @CountD >  0   THEN 4
         WHEN @CountN >  0   THEN 5
         WHEN @CountP >  0   THEN 6  
         WHEN @CountN1 >  0   THEN 7
         WHEN @CountN2 >  0   THEN 8
         WHEN @CountN3 >  0   THEN 9
         WHEN @CountN4 >  0   THEN 10
         WHEN @CountN5 >  0   THEN 11
         WHEN @CountN6 >  0   THEN 12   
         ELSE 1    
      END

RETURN @Zone  

END

它工作得很好，但是当每个表的数据都非常大时，就会因为超时而挂起。我想知道是否可以简化这个存储过程

这个存储过程非常简单：我只需输入

@TheInsee

，当在中找到它时，它会在六个表中搜索

---

有什么帮助吗？

首先，因为您只检查每个表中是否至少存在一行，所以计算所有行没有意义。其次，如果已经在上一个表中找到行，那么查询所有表是没有意义的

我会这样做：

set @Zone = case
    when exists (select * from dbo.LS_CLIENT_INSEE_A where NO_INSEE = @TheInsee) then 1
    when exists (select * from dbo.LS_CLIENT_INSEE_B where NO_INSEE = @TheInsee) then 2
    -- etc.
    else 1
    end

---

感谢您提供的解决方案当存在时我如何将select count（1）与select count结合起来？为什么

select*

不使用任何列

Select 1

效率更高，如：

Select 1 from dbo.LS_CLIENT_INSEE_A其中NO_INSEE=@TheInsee

@JayWalker在

存在的
查询中
Select 1
的效率是SQL Server的神话。MSDN在这种情况下使用星号，请参阅此问题以获得更详细的讨论：