使用索引的误区之一：没有使用复合索引的前导列导致查询不使用索引

在oracle中，我们经常以为建立了索引，sql查询的时候就会如我们所希望的那样使用索引，事实上，oracle只会在一定条件下使用索引，这里我们总结数第一点：oracle会在条件中包含了前导列时使用索引，即查询条件中必须使用索引中的第一个列，请看下面的例子

SQL> select * from tab;

 

TNAME                          TABTYPE  CLUSTERID

------------------------------ ------- ----------

BONUS                          TABLE

DEPT                           TABLE

DUMMY                          TABLE

EMP                            TABLE

SALGRADE                       TABLE

 

建立一个联合索引（注意复合索引的索引列顺序）

SQL> create index emp_id1 on emp(empno,ename,deptno);

 

Index created

 

建立一个单键索引

SQL> create index emp_id2 on emp(sal);

 

Index created

 

 

SQL> select table_name,index_name from user_indexes

  2  where table_name='EMP';

 

TABLE_NAME                     INDEX_NAME

------------------------------ ------------------------------

EMP                            EMP_ID1

EMP                            EMP_ID2

 

SQL> SELECT * FROM USER_IND_COLUMNS

  2  /

 

INDEX_NAME                     TABLE_NAME                     COLUMN_NAME                                                                      COLUMN_POSITION COLUMN_LENGTH CHAR_LENGTH DESCEND

------------------------------ ------------------------------ -------------------------------------------------------------------------------- --------------- ------------- ----------- -------

EMP_ID1                        EMP                            EMPNO                                                                                          1            22           0 ASC

EMP_ID1                        EMP                            ENAME                                                                                          2            10          10 ASC

EMP_ID1                        EMP                            DEPTNO                                                                                         3            22           0 ASC

EMP_ID2                        EMP                            SAL                                                                                            1            22           0 ASC

 

下面的查询由于没有使用到复合索引的前导列，所以没有使用索引

 select job, empno from emp where ename='RICH';

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

--------------------------------------------------------------------

| Id  | Operation            |  Name       | Rows  | Bytes | Cost  |

--------------------------------------------------------------------

|   0 | SELECT STATEMENT     |             |       |       |       |

|*  1 |  TABLE ACCESS FULL   | EMP         |       |       |       |

--------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   1 - filter("EMP"."ENAME"='RICH')

Note: rule based optimization

 

14 rows selected

 

 

下面的查询也由于没有使用到复合索引的前导列，所以没有使用索引

select job, empno from emp where deptno=30;

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

--------------------------------------------------------------------

| Id  | Operation            |  Name       | Rows  | Bytes | Cost  |

--------------------------------------------------------------------

|   0 | SELECT STATEMENT     |             |       |       |       |

|*  1 |  TABLE ACCESS FULL   | EMP         |       |       |       |

--------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   1 - filter("EMP"."DEPTNO"=30)

Note: rule based optimization

 

14 rows selected

 

 

 

下面的查询使用了复合索引中的前导列，所以查询走索引了

select job, empno from emp where empno=7777;

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

---------------------------------------------------------------------------

| Id  | Operation                   |  Name       | Rows  | Bytes | Cost  |

---------------------------------------------------------------------------

|   0 | SELECT STATEMENT            |             |       |       |       |

|   1 |  TABLE ACCESS BY INDEX ROWID| EMP         |       |       |       |

|*  2 |   INDEX RANGE SCAN          | EMP_ID1     |       |       |       |

---------------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   2 - access("EMP"."EMPNO"=7777)

Note: rule based optimization

 

15 rows selected

 

 

 

 

下面的查询使用了复合索引中的第一列和第二列，所以查询走索引了

select job, empno from emp where empno=7777 and ename='RICH';

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

---------------------------------------------------------------------------

| Id  | Operation                   |  Name       | Rows  | Bytes | Cost  |

---------------------------------------------------------------------------

|   0 | SELECT STATEMENT            |             |       |       |       |

|   1 |  TABLE ACCESS BY INDEX ROWID| EMP         |       |       |       |

|*  2 |   INDEX RANGE SCAN          | EMP_ID1     |       |       |       |

---------------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   2 - access("EMP"."EMPNO"=7777 AND "EMP"."ENAME"='RICH')

Note: rule based optimization

 

15 rows selected

 

 

 

使用了复合索引的全部列，所以走索引了，另外由于选了了索引中没有包含的列（job），

所以进行索引全表扫描得到满足条件的rowid后，还要到表中检索相应的行

select job, empno from emp where empno=7777 and ename='RICH' and deptno=30;

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

---------------------------------------------------------------------------

| Id  | Operation                   |  Name       | Rows  | Bytes | Cost  |

---------------------------------------------------------------------------

|   0 | SELECT STATEMENT            |             |       |       |       |

|   1 |  TABLE ACCESS BY INDEX ROWID| EMP         |       |       |       |

|*  2 |   INDEX RANGE SCAN          | EMP_ID1     |       |       |       |

---------------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   2 - access("EMP"."EMPNO"=7777 AND "EMP"."ENAME"='RICH' AND "EMP"."DEP

              TNO"=30)

Note: rule based optimization

 

16 rows selected

 

 

 

 

使用了复合索引的全部列，所以走索引了，而且由于所有选择的列都包含在索引中，所以仅仅进行了索引范围扫描

select empno from emp where empno=7777 and ename='RICH' and deptno=30;

 

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------

--------------------------------------------------------------------

| Id  | Operation            |  Name       | Rows  | Bytes | Cost  |

--------------------------------------------------------------------

|   0 | SELECT STATEMENT     |             |       |       |       |

|*  1 |  INDEX RANGE SCAN    | EMP_ID1     |       |       |       |

--------------------------------------------------------------------

Predicate Information (identified by operation id):

---------------------------------------------------

   1 - access("EMP"."EMPNO"=7777 AND "EMP"."ENAME"='RICH' AND "EM

              P"."DEPTNO"=30)

Note: rule based optimization

 

15 rows selected