Oracle 11g 的PL/SQL函数结果缓存，oracle11g
模仿Oracle性能诊断艺术中的例子做了两个试验，书上说如果不用RELIES_ON，则函数依赖的对象发生的变更操作就不会导致结果缓存的失效操作(result_cache RELIES_ON(test1,test2))，试验证明不对，函数f1()并没有使用RELIES_ON，但表上的变化影响到了函数。
C:\Documents and Settings\guogang>sqlplus gg_test/gg_test@10.10.15.25_gg
SQL*Plus: Release 10.2.0.1.0 - Production on 星期一 8月 4 19:46:44 2014Copyright (c) 1982, 2005, Oracle. All rights reserved.
连接到:
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> select * from v$version;
BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
PL/SQL Release 11.2.0.1.0 - Production
CORE 11.2.0.1.0 Production
TNS for Linux: Version 11.2.0.1.0 - Production
NLSRTL Version 11.2.0.1.0 - Production
SQL> drop table test1 purge;
SQL> drop table test2 purge;
SQL> create table test1 as select * from dba_objects;
SQL> create table test2 as select * from all_objects;
SQL> select count(*) from test1;
COUNT(*)
----------
74144
SQL> select count(*) from test2;
COUNT(*)
----------
73248
SQL> create or replace function f1
return number
is
l_ret number;
begin
select count(*) into l_ret
from test1,test2
where test1.object_type = test2.object_type
and test1.object_type in ('TABLE SUBPARTITION','VIEW','INDEX','TABLE');
return l_ret;
end;
/
函数已创建。
SQL> set timing on
SQL> select f1() from dual;
F1()
----------
60681409
已用时间: 00: 00: 07.29
--禁用结果缓存
SQL> execute dbms_result_cache.Bypass(bypass_mode=>true,session=>true);SQL> select f1() from dual;
F1()
----------
60681409
已用时间: 00: 00: 03.60
--启用结果缓存
SQL> select f1() from dual;
F1()
----------
60681409
已用时间: 00: 00: 00.00
SQL> delete from test1 where object_type = 'VIEW' and rownum <100;
SQL> delete from test2 where object_type = 'VIEW' and rownum <100;
SQL> commit;
SQL> select f1() from dual;
F1()
----------
59788330
已用时间: 00: 00: 07.09 --可以看到数据发生变化，即使不使用RELIES_ON，结果集也是正确的。
from test1, test2
where test1.object_type = test2.object_type
and test1.object_type in ('TABLE SUBPARTITION','VIEW','INDEX','TABLE');
COUNT(*)
----------
59788330
已用时间: 00: 00: 03.56
SQL> create or replace function f2
return number
result_cache RELIES_ON(test1,test2)
is
l_ret number;
begin
select count(*) into l_ret
from test1,test2
where test1.object_type = test2.object_type
and test1.object_type in ('TABLE SUBPARTITION','VIEW','INDEX','TABLE');
return l_ret;
end;
/
函数已创建。
SQL> select f2() from dual;
F2()
----------
59788330
已用时间: 00: 00: 03.54
SQL> select f2() from dual;
F2()
----------
59788330
已用时间: 00: 00: 00.00
SQL> delete from test1 where object_type = 'VIEW' and rownum <100;
SQL> delete from test2 where object_type = 'VIEW' and rownum <100;
SQL> commit;
SQL> select f2() from dual;
F2()
----------
58914853
已用时间: 00: 00: 03.50
from test1, test2
where test1.object_type = test2.object_type
and test1.object_type in ('TABLE SUBPARTITION','VIEW','INDEX','TABLE');
COUNT(*)
----------
58914853
已用时间: 00: 00: 03.50
如果预显示更新后结果，执行下面语句：
select deptno,empno,sal,decode(deptno,10,'5%',20,'10%',30,'15%') "sal%",
decode(deptno,10,sal*1.05,20,sal*1.1,30,sal*1.15) newsal
from emp order by deptno,empno;
若要先更新再显示，需要在表结构中增加更新后的列newsal。否则无法输出更新前的薪金。
更新：update emp x newset sal=(select decode(deptno,10,sal*1.05,20,sal*1.1,30,sal*1.15)
from emp where empno=x.empno);
显示：select deptno,empno,sal,decode(deptno,10,'5%',20,'10%',30,'15%') "sal%",newsal
from emp order by deptno,empno;
每次查询前清空缓存，试试：
10g以上：
alter system flush buffer_cache;
9i：
ALTER SESSION SET EVENTS 'immediate trace name flush_cache';