用解释计划评估创建索引后对单表查询效率的影响

有这么一张表：

create table hy_emp(
   empno number(6,0) primary key,
   deptno number(6,0) not null,
   ename nvarchar2(20) not null
)

这么给它充值：

insert into hy_emp
select rownum,dbms_random.value(1,12),dbms_random.string('*',dbms_random.value(6,20))
from dual 
connect by level<500000
order by dbms_random.random

下面我们对ename字段进行查询并取解释计划查看效率：

EXPLAIN PLAN FOR
select * from hy_emp where ename='OODFSK'

select * from table(dbms_xplan.display);

得到解释计划如下：

Plan hash value: 910676026
 
----------------------------------------------------------------------------
| Id  | Operation         | Name   | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------
|   0 | SELECT STATEMENT  |        |    30 |  1440 |   754   (1)| 00:00:01 |
|*  1 |  TABLE ACCESS FULL| HY_EMP |    30 |  1440 |   754   (1)| 00:00:01 |
----------------------------------------------------------------------------
 
Predicate Information (identified by operation id):
---------------------------------------------------
 
   1 - filter("ENAME"=U'OODFSK')
 
Note
-----
   - dynamic statistics used: dynamic sampling (level=2)

Cost是754。然后我们在ename字段上添加索引：

create index hy_emp_ename on hy_emp(ename);

再次查看解释计划：

EXPLAIN PLAN FOR
select * from hy_emp where ename='OODFSK'

select * from table(dbms_xplan.display);

解释计划如下：

Plan hash value: 2159543048
 
----------------------------------------------------------------------------------------------------
| Id  | Operation                           | Name         | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                    |              |     1 |    48 |     4   (0)| 00:00:01 |
|   1 |  TABLE ACCESS BY INDEX ROWID BATCHED| HY_EMP       |     1 |    48 |     4   (0)| 00:00:01 |
|*  2 |   INDEX RANGE SCAN                  | HY_EMP_ENAME |     1 |       |     3   (0)| 00:00:01 |
----------------------------------------------------------------------------------------------------
 
Predicate Information (identified by operation id):
---------------------------------------------------
 
   2 - access("ENAME"=U'OODFSK')
 
Note
-----
   - dynamic statistics used: dynamic sampling (level=2)

Cost一下子从754变到4了，只剩下零头了，看来加索引效果显著。

然后我们删掉这个索引：

drop index hy_emp_ename

再对ename和deptno两个字段进行联合查询并查看执行计划：

EXPLAIN PLAN FOR
select * from hy_emp where ename='OODFSK' and deptno=9
select * from table(dbms_xplan.display);

解释计划如下：

Plan hash value: 910676026
 
----------------------------------------------------------------------------
| Id  | Operation         | Name   | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------
|   0 | SELECT STATEMENT  |        |    55 |  2640 |   754   (1)| 00:00:01 |
|*  1 |  TABLE ACCESS FULL| HY_EMP |    55 |  2640 |   754   (1)| 00:00:01 |
----------------------------------------------------------------------------
 
Predicate Information (identified by operation id):
---------------------------------------------------
 
   1 - filter("ENAME"=U'OODFSK' AND "DEPTNO"=9)
 
Note
-----
   - dynamic statistics used: dynamic sampling (level=2)

Cost是754.

再给ename和deptno字段添加索引：

create index hy_emp_two on hy_emp(ename,deptno);

然后再看解释计划：

EXPLAIN PLAN FOR
select * from hy_emp where ename='OODFSK' and deptno=9
select * from table(dbms_xplan.display);

解释计划如下：

Plan hash value: 3007671827
 
--------------------------------------------------------------------------------------------------
| Id  | Operation                           | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
--------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                    |            |     1 |    48 |     4   (0)| 00:00:01 |
|   1 |  TABLE ACCESS BY INDEX ROWID BATCHED| HY_EMP     |     1 |    48 |     4   (0)| 00:00:01 |
|*  2 |   INDEX RANGE SCAN                  | HY_EMP_TWO |     1 |       |     3   (0)| 00:00:01 |
--------------------------------------------------------------------------------------------------
 
Predicate Information (identified by operation id):
---------------------------------------------------
 
   2 - access("ENAME"=U'OODFSK' AND "DEPTNO"=9)
 
Note
-----
   - dynamic statistics used: dynamic sampling (level=2)

又从754变成4了。

结论：索引对于提升查询效率明显，但前提是要和查询条件对应上，查几个字段就为这几个字段建立索引最好。

--2020-03-17--