exec sp_executesql @sql和exec（@sql）SQL Server

A Dynamic SQL query from lobodava is:

lobodava的动态SQL查询是：

declare @sql nvarchar(4000) =
    N';with cteColumnts (ORDINAL_POSITION, COLUMN_NAME) as 
    (
        select ORDINAL_POSITION, COLUMN_NAME 
        from INFORMATION_SCHEMA.COLUMNS 
        where TABLE_NAME = N'''+ @tableName + ''' and COLUMN_NAME like ''' + @columnLikeFilter + '''
    ),
    cteValues (ColumnName, SumValue) as
    (
        SELECT ColumnName, SumValue
        FROM 
           (SELECT ' + @sumColumns + '
           FROM dbo.' + @tableName + ') p
        UNPIVOT
           (SumValue FOR ColumnName IN 
              (' + @columns + ')
        )AS unpvt 
    )
    select row_number() over(order by ORDINAL_POSITION) as ID, ColumnName, SumValue
    from cteColumnts c inner join cteValues v on COLUMN_NAME = ColumnName
    order by ORDINAL_POSITION'

exec sp_executesql @sql
--OR
exec (@sql)

exec sp_executesql @sql --OR exec（@sql）

Why did lobodava pick exec sp_executesql @sql and not exec(@sql) So what is the difference here?
Is it better to use sp_executesql on recursive dynamic queries?
In other post they say sp_executesql is more likely to promote query plan reuse... So it helps in these kind of queries?

为什么lobodava选择exec sp_executesql @sql而不是exec（@sql）那么这里有什么区别？在递归动态查询中使用sp_executesql更好吗？在其他帖子中，他们说sp_executesql更有可能促进查询计划重用...所以它有助于这些类型的查询？

1 个解决方案

#1

Because EXEC sp_executesql will cache the query plan -- EXEC will not. For more info, and a very good read, see:

因为EXEC sp_executesql将缓存查询计划 - EXEC不会。有关更多信息和非常好的阅读，请参阅：

The Curse and Blessings of Dynamic SQL
动态SQL的诅咒与祝福

Caching a query means that the logistics to the query are temporarily stored, and make running the query later on faster for it.

缓存查询意味着临时存储查询的后勤，并使以后更快地运行查询。

#1