一个脚本来管理它们:这个脚本将截断和删除命令与sp_MSforeachtable结合在一起，这样您就可以避免删除和重新创建约束——只需指定需要删除而不是截断的表，为了便于度量，我还包含了一个额外的模式过滤器(在2008r2中测试)。

declare @schema nvarchar(max) = 'and Schema_Id=Schema_id(''Value'')'
declare @deletiontables nvarchar(max) = '(''TableA'',''TableB'')'
declare @truncateclause nvarchar(max) = @schema + ' and o.Name not in ' +  + @deletiontables;
declare @deleteclause nvarchar(max) = @schema + ' and o.Name in ' + @deletiontables;        

exec sp_MSforeachtable 'alter table ? nocheck constraint all', @whereand=@schema
exec sp_MSforeachtable 'truncate table ?', @whereand=@truncateclause
exec sp_MSforeachtable 'delete from ?', @whereand=@deleteclause
exec sp_MSforeachtable 'alter table ? with check check constraint all', @whereand=@schema

   --Drop and Recreate Foreign Key Constraints

SET NOCOUNT ON

DECLARE @table TABLE(
   RowId INT PRIMARY KEY IDENTITY(1, 1),
   ForeignKeyConstraintName NVARCHAR(200),
   ForeignKeyConstraintTableSchema NVARCHAR(200),
   ForeignKeyConstraintTableName NVARCHAR(200),
   ForeignKeyConstraintColumnName NVARCHAR(200),
   PrimaryKeyConstraintName NVARCHAR(200),
   PrimaryKeyConstraintTableSchema NVARCHAR(200),
   PrimaryKeyConstraintTableName NVARCHAR(200),
   PrimaryKeyConstraintColumnName NVARCHAR(200)    
)

INSERT INTO @table(ForeignKeyConstraintName, ForeignKeyConstraintTableSchema, ForeignKeyConstraintTableName, ForeignKeyConstraintColumnName)
SELECT 
   U.CONSTRAINT_NAME, 
   U.TABLE_SCHEMA, 
   U.TABLE_NAME, 
   U.COLUMN_NAME 
FROM 
   INFORMATION_SCHEMA.KEY_COLUMN_USAGE U
      INNER JOIN INFORMATION_SCHEMA.TABLE_CONSTRAINTS C
         ON U.CONSTRAINT_NAME = C.CONSTRAINT_NAME
WHERE
   C.CONSTRAINT_TYPE = 'FOREIGN KEY'

UPDATE @table SET
   PrimaryKeyConstraintName = UNIQUE_CONSTRAINT_NAME
FROM 
   @table T
      INNER JOIN INFORMATION_SCHEMA.REFERENTIAL_CONSTRAINTS R
         ON T.ForeignKeyConstraintName = R.CONSTRAINT_NAME

UPDATE @table SET
   PrimaryKeyConstraintTableSchema  = TABLE_SCHEMA,
   PrimaryKeyConstraintTableName  = TABLE_NAME
FROM @table T
   INNER JOIN INFORMATION_SCHEMA.TABLE_CONSTRAINTS C
      ON T.PrimaryKeyConstraintName = C.CONSTRAINT_NAME

UPDATE @table SET
   PrimaryKeyConstraintColumnName = COLUMN_NAME
FROM @table T
   INNER JOIN INFORMATION_SCHEMA.KEY_COLUMN_USAGE U
      ON T.PrimaryKeyConstraintName = U.CONSTRAINT_NAME

--SELECT * FROM @table

--DROP CONSTRAINT:
SELECT
   '
   ALTER TABLE [' + ForeignKeyConstraintTableSchema + '].[' + ForeignKeyConstraintTableName + '] 
   DROP CONSTRAINT ' + ForeignKeyConstraintName + '

   GO'
FROM
   @table

--ADD CONSTRAINT:
SELECT
   '
   ALTER TABLE [' + ForeignKeyConstraintTableSchema + '].[' + ForeignKeyConstraintTableName + '] 
   ADD CONSTRAINT ' + ForeignKeyConstraintName + ' FOREIGN KEY(' + ForeignKeyConstraintColumnName + ') REFERENCES [' + PrimaryKeyConstraintTableSchema + '].[' + PrimaryKeyConstraintTableName + '](' + PrimaryKeyConstraintColumnName + ')

   GO'
FROM
   @table

GO

我同意你的观点，哈姆林。当您使用SSIS传输数据或想要复制数据时，似乎很有必要暂时禁用或删除外键约束，然后重新启用或重新创建它们。在这些情况下，引用完整性不是问题，因为它已经在源数据库中进行了维护。所以，这件事你可以放心。

SQL-92标准允许将约束声明为DEFERRABLE，以便在事务范围内(隐式或显式)延迟它。遗憾的是，SQL Server仍然缺少这个SQL-92功能。

对我来说，将约束更改为NOCHECK类似于动态更改数据库结构——删除约束当然是这样——这是需要避免的(例如，用户需要增加特权)。

几乎可以肯定，检查是必须的!

在一些回答和评论中提出了这一点，但我认为再次提出这一点非常重要。

使用以下命令(不使用WITH CHECK)重新启用约束会有一些严重的缺陷。

ALTER TABLE MyTable CHECK CONSTRAINT MyConstraint;

WITH CHECK | WITH NOCHECK Specifies whether the data in the table is or is not validated against a newly added or re-enabled FOREIGN KEY or CHECK constraint. If not specified, WITH CHECK is assumed for new constraints, and WITH NOCHECK is assumed for re-enabled constraints. If you do not want to verify new CHECK or FOREIGN KEY constraints against existing data, use WITH NOCHECK. We do not recommend doing this, except in rare cases. The new constraint will be evaluated in all later data updates. Any constraint violations that are suppressed by WITH NOCHECK when the constraint is added may cause future updates to fail if they update rows with data that does not comply with the constraint. The query optimizer does not consider constraints that are defined WITH NOCHECK. Such constraints are ignored until they are re-enabled by using ALTER TABLE table WITH CHECK CHECK CONSTRAINT ALL.

注意:WITH NOCHECK是重新启用约束的默认值。我想知道为什么……

No existing data in the table will be evaluated during the execution of this command - successful completion is no guarantee that the data in the table is valid according to the constraint. During the next update of the invalid records, the constraint will be evaluated and will fail - resulting in errors that may be unrelated to the actual update that is made. Application logic that relies on the constraint to ensure that data is valid may fail. The query optimizer will not make use of any constraint that is enabled in this way.

sys。Foreign_keys系统视图提供了一些问题的可见性。注意，它有一个is_disabled列和一个is_not_trusted列。Is_disabled指示是否根据约束验证未来的数据操作。Is_not_trusted指示表中当前的所有数据是否已经根据约束进行了验证。

ALTER TABLE MyTable WITH CHECK CHECK CONSTRAINT MyConstraint;

你的约束是可信的吗?找出……

SELECT * FROM sys.foreign_keys WHERE is_not_trusted = 1;

一个脚本来管理它们:这个脚本将截断和删除命令与sp_MSforeachtable结合在一起，这样您就可以避免删除和重新创建约束——只需指定需要删除而不是截断的表，为了便于度量，我还包含了一个额外的模式过滤器(在2008r2中测试)。

declare @schema nvarchar(max) = 'and Schema_Id=Schema_id(''Value'')'
declare @deletiontables nvarchar(max) = '(''TableA'',''TableB'')'
declare @truncateclause nvarchar(max) = @schema + ' and o.Name not in ' +  + @deletiontables;
declare @deleteclause nvarchar(max) = @schema + ' and o.Name in ' + @deletiontables;        

exec sp_MSforeachtable 'alter table ? nocheck constraint all', @whereand=@schema
exec sp_MSforeachtable 'truncate table ?', @whereand=@truncateclause
exec sp_MSforeachtable 'delete from ?', @whereand=@deleteclause
exec sp_MSforeachtable 'alter table ? with check check constraint all', @whereand=@schema

最好的选择是DROP和CREATE外键约束。

我在这篇文章中没有找到对我“原样”工作的例子，如果外键引用不同的模式，一个不会工作，如果外键引用多列，另一个也不会工作。这个脚本同时考虑多个模式和每个外键的多个列。

下面是生成“ADD CONSTRAINT”语句的脚本，对于多列，它将用逗号分隔它们(确保在执行DROP语句之前保存此输出):

PRINT N'-- CREATE FOREIGN KEY CONSTRAINTS --';

SET NOCOUNT ON;
SELECT '
PRINT N''Creating '+ const.const_name +'...''
GO
ALTER TABLE ' + const.parent_obj + '
    ADD CONSTRAINT ' + const.const_name + ' FOREIGN KEY (
            ' + const.parent_col_csv + '
            ) REFERENCES ' + const.ref_obj + '(' + const.ref_col_csv + ')
GO'
FROM (
    SELECT QUOTENAME(fk.NAME) AS [const_name]
        ,QUOTENAME(schParent.NAME) + '.' + QUOTENAME(OBJECT_name(fkc.parent_object_id)) AS [parent_obj]
        ,STUFF((
                SELECT ',' + QUOTENAME(COL_NAME(fcP.parent_object_id, fcp.parent_column_id))
                FROM sys.foreign_key_columns AS fcP
                WHERE fcp.constraint_object_id = fk.object_id
                FOR XML path('')
                ), 1, 1, '') AS [parent_col_csv]
        ,QUOTENAME(schRef.NAME) + '.' + QUOTENAME(OBJECT_NAME(fkc.referenced_object_id)) AS [ref_obj]
        ,STUFF((
                SELECT ',' + QUOTENAME(COL_NAME(fcR.referenced_object_id, fcR.referenced_column_id))
                FROM sys.foreign_key_columns AS fcR
                WHERE fcR.constraint_object_id = fk.object_id
                FOR XML path('')
                ), 1, 1, '') AS [ref_col_csv]
    FROM sys.foreign_key_columns AS fkc
    INNER JOIN sys.foreign_keys AS fk ON fk.object_id = fkc.constraint_object_id
    INNER JOIN sys.objects AS oParent ON oParent.object_id = fkc.parent_object_id
    INNER JOIN sys.schemas AS schParent ON schParent.schema_id = oParent.schema_id
    INNER JOIN sys.objects AS oRef ON oRef.object_id = fkc.referenced_object_id
    INNER JOIN sys.schemas AS schRef ON schRef.schema_id = oRef.schema_id
    GROUP BY fkc.parent_object_id
        ,fkc.referenced_object_id
        ,fk.NAME
        ,fk.object_id
        ,schParent.NAME
        ,schRef.NAME
    ) AS const
ORDER BY const.const_name

下面是生成“DROP CONSTRAINT”语句的脚本:

PRINT N'-- DROP FOREIGN KEY CONSTRAINTS --';

SET NOCOUNT ON;

SELECT '
PRINT N''Dropping ' + fk.NAME + '...''
GO
ALTER TABLE [' + sch.NAME + '].[' + OBJECT_NAME(fk.parent_object_id) + ']' + ' DROP  CONSTRAINT ' + '[' + fk.NAME + ']
GO'
FROM sys.foreign_keys AS fk
INNER JOIN sys.schemas AS sch ON sch.schema_id = fk.schema_id
ORDER BY fk.NAME