xsw222 发表于 2018-10-17 07:35:05

SQL Server中表联接的执行计划比较

  1、运行如下语句:
  declare @t table( t1 int)
  declare @tt table(t1 int ,t2 varchar(10))
  insert into @t values(1)
  insert into @t values(2)
  insert into @t values(3)
  insert into @t values(4)
  insert into @tt values(1,'ab')
  insert into @tt values(1,'acd')
  insert into @tt values(2,'ab')
  insert into @tt values(3,'a')
  insert into @tt values(2,'a')
  set statistics profile on
  select *
  from @t x
  left outer join @tt y
  on x.t1 = y.t1
  where x.t1 = 1 and y.t2 like '%a%'
  结果:
  select *from @t xleft outer join @tt y    on x.t1 = y.t1   where x.t1 = 1 and y.t2 like '%a%'
  |--Nested Loops(Inner Join)
  |--Table Scan(OBJECT:(@tt AS ), WHERE:(@tt. as .=(1) AND @tt. as . like '%a%'))
  |--Table Scan(OBJECT:(@t AS ), WHERE:(@t. as .=(1)))
  2、把上面语句中的left outer join 改为inner join后
  结果:
  select *from @t xinner join @tt y    on x.t1 = y.t1   where x.t1 = 1 and y.t2 like '%a%'
  |--Nested Loops(Inner Join)
  |--Table Scan(OBJECT:(@tt AS ), WHERE:(@tt. as .=(1) AND @tt. as . like '%a%'))
  |--Table Scan(OBJECT:(@t AS ), WHERE:(@t. as .=(1)))
  3、如果过滤条件放到 on 子句里面,把上面的SQL改为:
  select *
  from @t x
  left outer join @tt y
  on x.t1 = y.t1   and    y.t2 like '%a%'
  wherex.t1 = 1
  产生的执行计划:
  select *from @t xleft outer join @tt y      on x.t1 = y.t1 and y.t2 like '%a%'where x.t1 = 1
  |--Nested Loops(Left Outer Join)
  |--Table Scan(OBJECT:(@t AS ), WHERE:(@t. as .=(1)))
  |--Table Scan(OBJECT:(@tt AS ), WHERE:(@tt. as .=(1) AND @tt. as . like '%a%'))
  其实对比上面的1、2、3,会发现在Table Scan里都是一样的 :
  |--Table Scan(OBJECT:(@t AS ), WHERE:(@t. as .=(1)))
  |--Table Scan(OBJECT:(@tt AS ), WHERE:(@tt. as .=(1) AND @tt. as . like '%a%'))
  而且在两个表关联之前,在Table Scan中已经用where子句过滤条件进行过滤,如果用到on子句,就会转化成where子句条件,
  只是对不同的情况,有不同的处理。
  用left outer join做关联时:
  一、当把右边表的字段过滤条件写在where中时(不用转化),在执行计划中是:Nested Loops(inner join)。
  这里SQL Server 做了优化(inner join的效率优于left outer join)。
  之所以可以做这样的优化,是由于把右边表的字段过滤条件直接放到where子句中,对联接产生的结果集合中左边表有,而右边表没有的记录,在联接结果集中此字段的值必定为NULL,那么通过这个字段的过滤,必定把NULL的记录都过滤掉了,此时效果等同于inner join,所以把外联接(left outer join)转化成内联接(inner join)。
  二、当把右边表的字段过滤条件写在on中时(把on转化成where子句),执行计划中是:Nested Loops(left outer join),无法优化。
  因为由left outer join产生的结果集与inner join产生的结果集是不同的,无法优化。

页: [1]
查看完整版本: SQL Server中表联接的执行计划比较