A join method is the mechanism for joining two row sources. Depending on the statistics, the optimizer chooses the method with the lowest estimated cost.
It pick one record from the smaller table and searches for the match in another table and returns the data when match is found.Then it picks the next row from smaller table and do the same comparison with other table.
NL join is chosen when one of the tables is small in size.
USE_NL & NO_USE_NL are used to influence the optimizer to use or not to use the NESTED LOOP JOIN.
SQL> SELECT /*+ USE_NL(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 4192419542
---------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
---------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 10 (0)| 00:00:01 |
| 1 | NESTED LOOPS | | 14 | 308 | 10 (0)| 00:00:01 |
| 2 | TABLE ACCESS FULL| DEPT | 4 | 52 | 3 (0)| 00:00:01 |
|* 3 | TABLE ACCESS FULL| EMP | 4 | 36 | 2 (0)| 00:00:01 |
---------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
3 - filter("EMP"."DEPTNO"="DEPT"."DEPTNO")
Hint Report (identified by operation id / Query Block Name / Object Alias):
Total hints for statement: 1 (U - Unused (1))
---------------------------------------------------------------------------
2 - SEL$1 / DEPT@SEL$1
U - USE_NL(emp,dept)
SQL> SELECT /*+ NO_USE_NL(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 844388907
----------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 6 (17)| 00:00:01 |
| 1 | MERGE JOIN | | 14 | 308 | 6 (17)| 00:00:01 |
| 2 | TABLE ACCESS BY INDEX ROWID| DEPT | 4 | 52 | 2 (0)| 00:00:01 |
| 3 | INDEX FULL SCAN | PK_DEPT | 4 | | 1 (0)| 00:00:01 |
|* 4 | SORT JOIN | | 14 | 126 | 4 (25)| 00:00:01 |
| 5 | TABLE ACCESS FULL | EMP | 14 | 126 | 3 (0)| 00:00:01 |
----------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
4 - access("EMP"."DEPTNO"="DEPT"."DEPTNO")
filter("EMP"."DEPTNO"="DEPT"."DEPTNO")
It picks the small table and then create a hash value based on the joining key column and placed the hash value in a table in memory.It then scans another table and compare the values with the in-memory hash table.A match is a join.The performance depends upon the size of join key and the available memory for the in-memory hash table.
It is only used in case of EQUI-JOIN.
USE_HASH & NO_USE_HASH are used to influence the optimizer to use or not to use the HASH JOIN.
SQL> SELECT /*+ USE_HASH(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 615168685
---------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
---------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 6 (0)| 00:00:01 |
|* 1 | HASH JOIN | | 14 | 308 | 6 (0)| 00:00:01 |
| 2 | TABLE ACCESS FULL| DEPT | 4 | 52 | 3 (0)| 00:00:01 |
| 3 | TABLE ACCESS FULL| EMP | 14 | 126 | 3 (0)| 00:00:01 |
---------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("EMP"."DEPTNO"="DEPT"."DEPTNO")
Hint Report (identified by operation id / Query Block Name / Object Alias):
Total hints for statement: 1 (U - Unused (1))
---------------------------------------------------------------------------
2 - SEL$1 / DEPT@SEL$1
U - USE_HASH(emp,dept)
SQL> SELECT /*+ USE_NO_HASH(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 844388907
----------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 6 (17)| 00:00:01 |
| 1 | MERGE JOIN | | 14 | 308 | 6 (17)| 00:00:01 |
| 2 | TABLE ACCESS BY INDEX ROWID| DEPT | 4 | 52 | 2 (0)| 00:00:01 |
| 3 | INDEX FULL SCAN | PK_DEPT | 4 | | 1 (0)| 00:00:01 |
|* 4 | SORT JOIN | | 14 | 126 | 4 (25)| 00:00:01 |
| 5 | TABLE ACCESS FULL | EMP | 14 | 126 | 3 (0)| 00:00:01 |
----------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
4 - access("EMP"."DEPTNO"="DEPT"."DEPTNO")
filter("EMP"."DEPTNO"="DEPT"."DEPTNO")
Hint Report (identified by operation id / Query Block Name / Object Alias):
Total hints for statement: 1 (E - Syntax error (1))
---------------------------------------------------------------------------
1 - SEL$1
E - USE_NO_HASH
Oracle sorts the first table by its join columns,sorts the second table by its join columns and then merges the sorted row source together.As matches are found they are put in into the result-set.This type of join is effective when there is a lack of data selectivity or NL join is insufficient or both the tables are quite large.
USE_MERGE & NO_USE_MERGE are used to influence the optimizer to use or not to use the SORT MERGE JOIN.
SQL> SELECT /*+ USE_MERGE(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 844388907
----------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 6 (17)| 00:00:01 |
| 1 | MERGE JOIN | | 14 | 308 | 6 (17)| 00:00:01 |
| 2 | TABLE ACCESS BY INDEX ROWID| DEPT | 4 | 52 | 2 (0)| 00:00:01 |
| 3 | INDEX FULL SCAN | PK_DEPT | 4 | | 1 (0)| 00:00:01 |
|* 4 | SORT JOIN | | 14 | 126 | 4 (25)| 00:00:01 |
| 5 | TABLE ACCESS FULL | EMP | 14 | 126 | 3 (0)| 00:00:01 |
----------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
4 - access("EMP"."DEPTNO"="DEPT"."DEPTNO")
filter("EMP"."DEPTNO"="DEPT"."DEPTNO")
Hint Report (identified by operation id / Query Block Name / Object Alias):
Total hints for statement: 1 (U - Unused (1))
---------------------------------------------------------------------------
2 - SEL$1 / DEPT@SEL$1
U - USE_MERGE(emp,dept)
SQL> SELECT /*+ NO_USE_MERGE(emp,dept) */ ename,dname FROM emp,dept WHERE emp.deptno = dept.deptno;
Execution Plan
----------------------------------------------------------
Plan hash value: 615168685
---------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
---------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 14 | 308 | 6 (0)| 00:00:01 |
|* 1 | HASH JOIN | | 14 | 308 | 6 (0)| 00:00:01 |
| 2 | TABLE ACCESS FULL| DEPT | 4 | 52 | 3 (0)| 00:00:01 |
| 3 | TABLE ACCESS FULL| EMP | 14 | 126 | 3 (0)| 00:00:01 |
---------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("EMP"."DEPTNO"="DEPT"."DEPTNO")