本文转载自微信公众号「DBA闲思杂想录」，作者潇湘隐者 。转载本文请联系DBA闲思杂想录公众号。
Oracle 11g在DBMS_SHARED_POOL包中引入了一个名为PURGE的新存储过程，用于从对象库缓存中刷新特定对象，例如游标，包，序列，触发器等。也就是说可以删除、清理特定SQL的执行计划， 这样在特殊情况下，就避免你要将整个SHARED POOL清空的危险情况。例如某个SQL语句由于优化器产生了错误的执行计划，我们希望优化器重新解析，生成新的执行计划，必须无将SQL的执行计划从共享池中刷出或将其置为无效，那么优化器才能将后续SQL进行硬解析、生成新的执行计划。这在以前只能使用清空共享池的方法或对表进行DDL操作。现在就可以指定刷新特定SQL的执行计划。当然在10.2.0.4 和10.2.0.5的补丁集中该包也被包含进来，该包的存储过程有三个参数，如下所示：
- DBMS_SHARED_POOL.PURGE (
- name VARCHAR2,
- flag CHAR DEFAULT 'P',
- heaps NUMBER DEFAULT 1)
- Argument Name Type In/Out Default?
- ------------------------------ ----------------------- ------ --------
- NAME VARCHAR2 IN
- FLAG CHAR IN DEFAULT
- HEAPS NUMBER IN DEFAULT
第一个参数为逗号分隔的ADDRESS列和HASH_VALUE列的值。
第二个参数可以有多个选项，例如C、P、T、R、Q等。具体意义如下所示 C表示PURGE的对象是CURSOR
- Set to 'P' or 'p' to fully specify that the input is the name of a package/procedure/function.
- Set to 'T' or 't' to specify that the input is the name of a type.
- Set to 'R' or 'r' to specify that the input is the name of a trigger.
- Set to 'Q' or 'q' to specify that the input is the name of a sequence.
- ...................................
第三个参数heaps，一般使用默认值1
- Heaps to be purged. For example, if heap 0 and heap 6 are to be purged:
- 1<<0 | 1<<6 => hex 0x41 => decimal 65, so specify heaps =>65.Default is 1, that is, heap 0 which means the whole object would be purged
在ORACLE 11g当中，你可以在$ORACLE_HOME/rdbms/admin/dbmspool.sql中查看该包的具体定义. 但是这个DBMS_SHARED_POOL.PURGE在10.2.0.4.0(实际测试发现10.2.0.5.0也存在同样问题)都有一些问题，它可能无法生效，当然在Oracle 11g中没有这个问题，具体演示如下所示：
- SQL> select * from v$version;
- BANNER
- ----------------------------------------------------------------
- Oracle Database 10g Release 10.2.0.5.0 - 64bit Production
- PL/SQL Release 10.2.0.5.0 - Production
- CORE 10.2.0.5.0 Production
- TNS for Linux: Version 10.2.0.5.0 - Production
- NLSRTL Version 10.2.0.5.0 - Production
- SQL> alter system flush shared_pool;
- System altered.
- SQL> set linesize 1200;
- SQL> select * from scott.dept where deptno=40;
- DEPTNO DNAME LOC
- ---------- -------------- -------------
- 40 OPERATIONS BOSTON
- SQL> select sql_id, first_load_time
- 2 from v$sql
- 3 where sql_text like 'select * from scott.dept%';
- SQL_ID FIRST_LOAD_TIME
- ------------- ---------------------------------------------------------
- 3nvuzqdn6ry6x 2016-12-29/08:51:21
- SQL> col sql_text for a64;
- SQL> select address, hash_value, sql_text
- 2 from v$sqlarea
- 3 where sql_id='3nvuzqdn6ry6x';
- ADDRESS HASH_VALUE SQL_TEXT
- ---------------- ---------- ----------------------------------------------------------------
- 00000000968ED510 1751906525 select * from scott.dept where deptno=40
- SQL> exec dbms_shared_pool.purge('00000000968ED510,1751906525','C');
- PL/SQL procedure successfully completed.
- SQL> select address, hash_value, sql_text
- 2 from v$sqlarea
- 3 where sql_id='3nvuzqdn6ry6x';
- ADDRESS HASH_VALUE SQL_TEXT
- ---------------- ---------- ----------------------------------------------------------------
- 00000000968ED510 1751906525 select * from scott.dept where deptno=40
- SQL>
如上截图所示，DBMS_SHARED_POOL.PURGE并没有清除这个特定的SQL的执行计划，其实这个是因为在10.2.0.4.0 要生效就必须开启5614566 EVNET，否则不会生效。具体可以参考官方文档：
- DBMS_SHARED_POOL.PURGE Is Not Working On 10.2.0.4 (文档 ID 751876.1)
- Bug 7538951 : DBMS_SHARED_POOL IS NOT WORKING AS EXPECTED
- Bug 5614566 : WE NEED A FLUSH CURSOR INTERFACE
- DBMS_SHARED_POOL.PURGE is available from 11.1. In 10.2.0.4, it is available
- through the fix for Bug 5614566. However, the fix is event protected. You need to set the event 5614566 to make use of purge. Unless the event is set, dbms_shared_pool.purge will have no effect.
- Set the event 5614566 in the init.ora to turn purge on.
- event="5614566 trace name context forever"
如下所示，设置5614566 event后，必须重启数据库才能生效，这个也是一个比较麻烦的事情。
- alter system set event = '5614566 trace name context forever' scope = spfile;