oracle-組合索引字段位置與查詢效率之間的關系

Oracle索引組合字段的位置不同，當查詢條件不能覆蓋索引時，影響查詢效率。查詢條件是不是索引字段的第一列影響執行計划，實驗驗證

實驗1：查詢條件為組合索引的第一列
--創建測試表
create table test_C
as
select * from user_objects where 1=2;

--創建測試表
create table test_a
as
select * from user_objects where 1=1;
--創建組合索引
create index test_a_1 on test_a(object_id,object_type);
set linesize 1000;
set autotrace trace;
select * from test_a where object_id=161728;
****************************************************
執行計划
----------------------------------------------------------
Plan hash value: 3036018411

------------------------------------------------------------------------------------------------
| Id  | Operation                           | Name     | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                    |          |     1 |   116 |     3   (0)| 00:00:01 |
|   1 |  TABLE ACCESS BY INDEX ROWID BATCHED| TEST_A   |     1 |   116 |     3   (0)| 00:00:01 |
|*  2 |   INDEX RANGE SCAN                  | TEST_A_1 |     1 |       |     2   (0)| 00:00:01 |
------------------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   2 - access("OBJECT_ID"=161728)


統計信息
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          3  consistent gets
          0  physical reads
          0  redo size
       1933  bytes sent via SQL*Net to client
        500  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          1  rows processed

查詢條件object_id為索引的前導列。執行計划：INDEX RANGE SCAN，cost是3，邏輯讀是3.

drop index test_a_1;
drop index test_a_2;
--創建索引2，查詢條件在索引中的位置是第二列
create index test_a_2 on test_a(object_name,object_ID);
--再次查詢
select * from test_a where object_id=161728;
****************************************************
執行計划
----------------------------------------------------------
Plan hash value: 219338936

------------------------------------------------------------------------------------------------
| Id  | Operation                           | Name     | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                    |          |     1 |   116 |     7   (0)| 00:00:01 |
|   1 |  TABLE ACCESS BY INDEX ROWID BATCHED| TEST_A   |     1 |   116 |     7   (0)| 00:00:01 |
|*  2 |   INDEX SKIP SCAN                   | TEST_A_2 |     1 |       |     6   (0)| 00:00:01 |
------------------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   2 - access("OBJECT_ID"=161728)
       filter("OBJECT_ID"=161728)


統計信息
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          8  consistent gets
          0  physical reads
          0  redo size
       1933  bytes sent via SQL*Net to client
        500  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          1  rows processed

查詢條件object_id為索引的第二列。執行計划：INDEX SKIP SCAN，cost是7，邏輯讀是8.

同樣object_id作為查詢條件，當object_id作為索引的第一列或者作為索引的第二列，它們在cost與邏輯讀上是有區別的，也就是索引跳躍掃描(INDEX SKIP SCAN)沒有索引范圍掃描(INDEX RANGE SCAN)的效率高。

組合索引字段的位置影響查詢效率，所以在建索引的時候需注意.

建議:把等值條件的字段放在組合索引的前導列,范圍字段放在第二列.