SQL Server，测试发现将条件中的@

SQL Server，测试发现将条件中的@

成本核算程序执行某个存储过程一直阻塞，排查发现类似以下语句阻塞：

select 
    tbl1.product_id,sum(isnull(tbl1.qty,0) * isnull(tbl2.unit_other_cost,0)) as other_cost
from tbl1.p_id=@pId and tbl1.if_stock=0
    and exists(select 1 from tbl3 inner join tbl4 on tbl3.c1=tbl4.c2 where tbl4.c3=0 and tbl3.p_id=tbl1.p_id and tbl3.product_id=tbl1.product_id)
group by tbl1.product_id

其中参数@pId是存储过程的传入参数。测试发现将条件中的@pId改成具体的值，直接执行SQL语句也会阻塞，但是加上变量定义就不会了：

declare @newpId int
set @newpId=99
select 
    tbl1.product_id,sum(isnull(tbl1.qty,0) * isnull(tbl2.unit_other_cost,0)) as other_cost
from tbl1.p_id=@newpId and tbl1.if_stock=0
    and exists(select 1 from tbl3 inner join tbl4 on tbl3.c1=tbl4.c2 where tbl4.c3=0 and tbl3.p_id=tbl1.p_id and tbl3.product_id=tbl1.product_id)
group by tbl1.product_id

最终决定存储过程里重新定义个变量，赋值为传入参数，将重新定义的变量作为条件值，问题解决。

declare @newpId int
set @newpId=@pId
select 
    tbl1.product_id,sum(isnull(tbl1.qty,0) * isnull(tbl2.unit_other_cost,0)) as other_cost
from tbl1.p_id=@newpId and tbl1.if_stock=0
    and exists(select 1 from tbl3 inner join tbl4 on tbl3.c1=tbl4.c2 where tbl4.c3=0 and tbl3.p_id=tbl1.p_id and tbl3.product_id=tbl1.product_id)
group by tbl1.product_id

同样的程序、存储过程其它工厂核算时没有问题，即其它数据库没有出现阻塞，只在这个数据库阻塞，可能跟数据库设置有关，原理需要再研究。