Non Unnestable Subquery

Subquery가 항상 Unnesting 되지는 않는다.
항상 실행 계획을 통해 Subquery Unnesting의 발생 여부를 확인해야 한다.

select /*+ gather_plan_statistics */
 t1.c1, t1.c2
from
  t1
where
   t1.c1 in (select c1 from t2) or
   t1.c1 between 1 and 100  
;

@stat
---------------------------------------------------------------------------------------------------
| Id  | Operation          | Name  | Starts | E-Rows | Cost (%CPU)| A-Rows |   A-Time   | Buffers |
---------------------------------------------------------------------------------------------------
|*  1 |  FILTER            |       |      1 |        |            |   1000 |00:00:00.03 |   19905 |
|   2 |   TABLE ACCESS FULL| T1    |      1 |  10000 |    11   (0)|  10000 |00:00:00.04 |     105 |
|*  3 |   INDEX RANGE SCAN | T2_N1 |   9900 |      1 |     1   (0)|    900 |00:00:00.12 |   19800 |
---------------------------------------------------------------------------------------------------

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

   1 - filter((("T1"."C1"<=100 AND "T1"."C1">=1) OR  IS NOT NULL))
   3 - access("C1"=:B1)

Oracle은 Or Operation이 사용되는 경우 Subquery Unnesting이 수행하지 못하는 한계를 지니고 있다.
이를 해결하는 유일한 방법은 Query를 재 작성하는 것이다.

select /*+ gather_plan_statistics */
  t1.c1, t1.c2
from
  t1
where
   t1.c1 in (select c1 from t2)
union
select 
  t1.c1, t1.c2
from
  t1
where
  t1.c1 between 1 and 100
;

@stat
-----------------------------------------------------------------------------------------------------------------------------------------
| Id  | Operation                     | Name  | Starts | E-Rows | Cost (%CPU)| A-Rows |   A-Time   | Buffers |  OMem |  1Mem | Used-Mem |
-----------------------------------------------------------------------------------------------------------------------------------------
|   1 |  SORT UNIQUE                  |       |      1 |   1100 |    19  (32)|   1000 |00:00:00.01 |   48 | 57344 | 57344 |51200  (0)|
|   2 |   UNION-ALL                   |       |      1 |        |            |   1100 |00:00:00.02 |   48 |          |       |          |
|*  3 |    HASH JOIN                  |       |      1 |   1000 |    14   (8)|   1000 |00:00:00.01 |   45 |     1517K|  1517K| 1466K (0)|
|   4 |     INDEX FAST FULL SCAN      | T2_N1 |      1 |   1000 |     2   (0)|   1000 |00:00:00.01 |    7 |          |       |          |
|   5 |     TABLE ACCESS FULL         | T1    |      1 |  10000 |    11   (0)|  10000 |00:00:00.04 |   38 |          |       |          |
|   6 |    TABLE ACCESS BY INDEX ROWID| T1    |      1 |    100 |     3   (0)|    100 |00:00:00.01 |    3 |          |       |          |
|*  7 |     INDEX RANGE SCAN          | T1_N1 |      1 |    100 |     2   (0)|    100 |00:00:00.01 |    2 |          |       |          |
-----------------------------------------------------------------------------------------------------------------------------------------

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

   3 - access("T1"."C1"="C1")
   7 - access("T1"."C1">=1 AND "T1"."C1"<=100)

or

select /*+ gather_plan_statistics  */
  t1.c1, t1.c2
from
  t1
where
   t1.c1 in (select c1 from t2 where t2.c1 not between 1 and 100)
union all
select 
  t1.c1, t1.c2
from
  t1
where
  t1.c1 between 1 and 100
;

@stat
----------------------------------------------------------------------------------------------------------------------------------------
| Id  | Operation                    | Name  | Starts | E-Rows | Cost (%CPU)| A-Rows |   A-Time   | Buffers |  OMem |  1Mem | Used-Mem |
----------------------------------------------------------------------------------------------------------------------------------------
|   1 |  UNION-ALL                   |       |      1 |        |            |   1000 |00:00:00.02 |  122 |          |       |          |
|*  2 |   HASH JOIN RIGHT SEMI       |       |      1 |    901 |    14   (8)|    900 |00:00:00.01 |  105 |  1517K|  1517K| 1465K (0)|
|*  3 |    INDEX FAST FULL SCAN      | T2_N1 |      1 |    901 |     2   (0)|    900 |00:00:00.01 |    7 |          |       |          |
|   4 |    TABLE ACCESS FULL         | T1    |      1 |  10000 |    11   (0)|  10000 |00:00:00.04 |   98 |          |       |          |
|   5 |   TABLE ACCESS BY INDEX ROWID| T1    |      1 |    100 |     3   (0)|    100 |00:00:00.01 |   17 |          |       |          |
|*  6 |    INDEX RANGE SCAN          | T1_N1 |      1 |    100 |     2   (0)|    100 |00:00:00.01 |    9 |          |       |          |
----------------------------------------------------------------------------------------------------------------------------------------

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

   2 - access("T1"."C1"="C1")
   3 - filter(("T2"."C1">100 OR "T2"."C1"<1))
   6 - access("T1"."C1">=1 AND "T1"."C1"<=100)

문서에 대하여

• 최초작성자 : 김종원
• 최초작성일 : 2009년 3월 21일
• 이 문서는 오라클클럽 코어 오라클 데이터베이스 스터디 모임에서 작성하였습니다.
• {*}이 문서의 내용은 조동욱님의 'Optimizing Oracle Optimizer'을 참고하였습니다.*