在谓词推送中，优化器将包含在查询块中的相关谓词推送到视图查询块中。对于不能合并的视图，这种技术可以提高不能合并视图的执行计划。数据库可以使用推送谓词来访问索引或作为过滤。
例如，假设创建了一个hr.contract_workers表:
SQL> drop table contract_workers; Table dropped. SQL> create table contract_workers as (select * from employees where 1=2); Table created. SQL> insert into contract_workers values (306, 'bill', 'jones', 'bjones','555.555.2000', '07-jun-02', 'ac_account', 8300, 0,205, 110); 1 row created. SQL> insert into contract_workers values (406, 'jill', 'ashworth', 'jashworth','555.999.8181', '09-jun-05', 'ac_account', 8300, 0,205, 50); 1 row created. SQL> insert into contract_workers values (506, 'marcie', 'lunsford', 'mlunsford','555.888.2233', '22-jul-01', 'ac_account', 8300, 0,205, 110); 1 row created. SQL> commit; Commit complete. SQL> create index contract_workers_index on contract_workers(department_id); Index created.
创建一个视图引用employees与contract_workers表。视图使用了union集合操作:
SQL> create view all_employees_vw as 2 select employee_id, last_name, job_id, commission_pct, department_id 3 from employees 4 union 5 select employee_id, last_name, job_id, commission_pct, department_id 6 from contract_workers; View created.
然后对视图执行查询:
select last_name from all_employees_vw where department_id = 50;
因为视图是一个union集合操作查询，优化器不能合并视图的查询到主查询块。优化器可以通过推送谓词来转换查询，where子句条件department_id=50，会推送到视图的union集合操作查询中，转换后的等价查询如下:
select last_name from ( select employee_id, last_name, job_id, commission_pct, department_id from employees where department_id=50 union select employee_id, last_name, job_id, commission_pct, department_id from contract_workers where department_id=50 );
转换后的查询现在可以考虑对每个查询块使用索引或全表扫描,查询视图语句的执行计划如下:
SQL> select * from table(dbms_xplan.display_cursor(null,null,'advanced allstats last runstats_last peeked_binds'));
SQL_ID 265ccrp674n30, child number 0
-------------------------------------
select last_name from all_employees_vw where department_id = 50
Plan hash value: 1422200799
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Starts | E-Rows |E-Bytes|E-Temp | Cost (%CPU)| E-Time | A-Rows | A-Time | Buffers | Reads | OMem | 1Mem | Used-Mem |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | | | | 1018 (100)| | 100K|00:00:01.37 | 955 | 942 | | | |
| 1 | VIEW | ALL_EMPLOYEES_VW | 1 | 100K| 2637K| | 1018 (1)| 00:00:01 | 100K|00:00:01.37 | 955 | 942 | | | |
| 2 | SORT UNIQUE | | 1 | 100K| 2540K| 3936K| 1018 (1)| 00:00:01 | 100K|00:00:01.18 | 955 | 942 | 8416K| 1135K| 7480K (0)|
| 3 | UNION-ALL | | 1 | | | | | | 100K|00:00:00.76 | 955 | 942 | | | |
|* 4 | TABLE ACCESS FULL| EMPLOYEES | 1 | 100K| 2540K| | 273 (1)| 00:00:01 | 100K|00:00:00.41 | 948 | 942 | | | |
|* 5 | TABLE ACCESS FULL| CONTRACT_WORKERS | 1 | 1 | 60 | | 2 (0)| 00:00:01 | 1 |00:00:00.01 | 7 | 0 | | | |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Query Block Name / Object Alias (identified by operation id):
-------------------------------------------------------------
1 - SET$1 / ALL_EMPLOYEES_VW@SEL$1
2 - SET$1
4 - SEL$2 / EMPLOYEES@SEL$2
5 - SEL$3 / CONTRACT_WORKERS@SEL$3
Outline Data
-------------
/*+
BEGIN_OUTLINE_DATA
IGNORE_OPTIM_EMBEDDED_HINTS
OPTIMIZER_FEATURES_ENABLE('12.2.0.1')
DB_VERSION('12.2.0.1')
ALL_ROWS
NO_PARALLEL
OUTLINE_LEAF(@"SEL$2")
OUTLINE_LEAF(@"SEL$3")
OUTLINE_LEAF(@"SET$1")
OUTLINE_LEAF(@"SEL$1")
NO_ACCESS(@"SEL$1" "ALL_EMPLOYEES_VW"@"SEL$1")
FULL(@"SEL$3" "CONTRACT_WORKERS"@"SEL$3")
FULL(@"SEL$2" "EMPLOYEES"@"SEL$2")
END_OUTLINE_DATA
*/
Predicate Information (identified by operation id):
---------------------------------------------------
4 - filter("DEPARTMENT_ID"=50)
5 - filter("DEPARTMENT_ID"=50)
从执行计划的Predicate Information部分可以看到4,5操作使用了department_id=50来分别对表employees和contract_workers来进行过滤，也证明了可以将谓词推送到了视图中的查询块。